| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.852366398073539, |
| "eval_steps": 500, |
| "global_step": 20000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.004630915995183847, |
| "grad_norm": 51.771697998046875, |
| "learning_rate": 1.5123456790123457e-07, |
| "loss": 4.9315, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.009261831990367694, |
| "grad_norm": 48.02553939819336, |
| "learning_rate": 3.055555555555556e-07, |
| "loss": 4.8131, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.013892747985551542, |
| "grad_norm": 50.21088409423828, |
| "learning_rate": 4.5987654320987657e-07, |
| "loss": 4.148, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.018523663980735388, |
| "grad_norm": 31.535659790039062, |
| "learning_rate": 6.141975308641976e-07, |
| "loss": 3.4211, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.023154579975919236, |
| "grad_norm": 24.023160934448242, |
| "learning_rate": 7.685185185185186e-07, |
| "loss": 2.6902, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.027785495971103084, |
| "grad_norm": 20.243444442749023, |
| "learning_rate": 9.228395061728395e-07, |
| "loss": 2.3382, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.03241641196628693, |
| "grad_norm": 17.97671890258789, |
| "learning_rate": 1.0771604938271605e-06, |
| "loss": 2.1599, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.037047327961470776, |
| "grad_norm": 18.1010684967041, |
| "learning_rate": 1.2314814814814815e-06, |
| "loss": 1.9409, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.04167824395665463, |
| "grad_norm": 16.637313842773438, |
| "learning_rate": 1.3858024691358028e-06, |
| "loss": 1.9179, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.04630915995183847, |
| "grad_norm": 22.70376205444336, |
| "learning_rate": 1.5401234567901236e-06, |
| "loss": 1.8012, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.05094007594702232, |
| "grad_norm": 20.450454711914062, |
| "learning_rate": 1.6944444444444446e-06, |
| "loss": 1.6436, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.05557099194220617, |
| "grad_norm": 16.855588912963867, |
| "learning_rate": 1.8487654320987655e-06, |
| "loss": 1.754, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.06020190793739002, |
| "grad_norm": 15.478442192077637, |
| "learning_rate": 2.0030864197530867e-06, |
| "loss": 1.6117, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.06483282393257386, |
| "grad_norm": 14.401535034179688, |
| "learning_rate": 2.1574074074074073e-06, |
| "loss": 1.574, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.06946373992775771, |
| "grad_norm": 19.8779296875, |
| "learning_rate": 2.311728395061729e-06, |
| "loss": 1.4455, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.07409465592294155, |
| "grad_norm": 23.51628875732422, |
| "learning_rate": 2.4660493827160494e-06, |
| "loss": 1.5821, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.0787255719181254, |
| "grad_norm": 13.794211387634277, |
| "learning_rate": 2.6203703703703705e-06, |
| "loss": 1.3384, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.08335648791330925, |
| "grad_norm": 17.96605110168457, |
| "learning_rate": 2.7746913580246915e-06, |
| "loss": 1.3484, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.0879874039084931, |
| "grad_norm": 18.876876831054688, |
| "learning_rate": 2.9290123456790126e-06, |
| "loss": 1.3963, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.09261831990367694, |
| "grad_norm": 21.498750686645508, |
| "learning_rate": 3.0833333333333336e-06, |
| "loss": 1.3247, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.0972492358988608, |
| "grad_norm": 15.515962600708008, |
| "learning_rate": 3.2376543209876542e-06, |
| "loss": 1.2645, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.10188015189404465, |
| "grad_norm": 19.241085052490234, |
| "learning_rate": 3.3919753086419753e-06, |
| "loss": 1.2757, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.10651106788922848, |
| "grad_norm": 19.796600341796875, |
| "learning_rate": 3.5462962962962967e-06, |
| "loss": 1.2339, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.11114198388441233, |
| "grad_norm": 818.7846069335938, |
| "learning_rate": 3.7006172839506178e-06, |
| "loss": 1.2809, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.11577289987959619, |
| "grad_norm": 13.61767578125, |
| "learning_rate": 3.854938271604939e-06, |
| "loss": 1.185, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.12040381587478004, |
| "grad_norm": 17.602561950683594, |
| "learning_rate": 4.0092592592592594e-06, |
| "loss": 1.1874, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.1250347318699639, |
| "grad_norm": 17.445096969604492, |
| "learning_rate": 4.163580246913581e-06, |
| "loss": 1.273, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.12966564786514773, |
| "grad_norm": 18.69367790222168, |
| "learning_rate": 4.3179012345679015e-06, |
| "loss": 1.1443, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.13429656386033156, |
| "grad_norm": 16.09065055847168, |
| "learning_rate": 4.472222222222223e-06, |
| "loss": 1.2184, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.13892747985551543, |
| "grad_norm": 16.429534912109375, |
| "learning_rate": 4.626543209876544e-06, |
| "loss": 1.1603, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.14355839585069927, |
| "grad_norm": 14.309248924255371, |
| "learning_rate": 4.780864197530864e-06, |
| "loss": 1.1405, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.1481893118458831, |
| "grad_norm": 15.651091575622559, |
| "learning_rate": 4.935185185185186e-06, |
| "loss": 1.1124, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.15282022784106697, |
| "grad_norm": 15.110349655151367, |
| "learning_rate": 5.089506172839507e-06, |
| "loss": 1.1964, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.1574511438362508, |
| "grad_norm": 14.926290512084961, |
| "learning_rate": 5.243827160493827e-06, |
| "loss": 1.0599, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.16208205983143464, |
| "grad_norm": 14.762713432312012, |
| "learning_rate": 5.398148148148149e-06, |
| "loss": 1.0663, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.1667129758266185, |
| "grad_norm": 13.759990692138672, |
| "learning_rate": 5.55246913580247e-06, |
| "loss": 1.0532, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.17134389182180235, |
| "grad_norm": 12.649763107299805, |
| "learning_rate": 5.7067901234567905e-06, |
| "loss": 1.1887, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.1759748078169862, |
| "grad_norm": 12.257176399230957, |
| "learning_rate": 5.861111111111112e-06, |
| "loss": 1.1441, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.18060572381217005, |
| "grad_norm": 13.522123336791992, |
| "learning_rate": 6.0154320987654326e-06, |
| "loss": 1.0646, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.1852366398073539, |
| "grad_norm": 13.428495407104492, |
| "learning_rate": 6.169753086419754e-06, |
| "loss": 1.0759, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.18986755580253775, |
| "grad_norm": 15.00213623046875, |
| "learning_rate": 6.324074074074075e-06, |
| "loss": 1.1082, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.1944984717977216, |
| "grad_norm": 14.586730003356934, |
| "learning_rate": 6.478395061728395e-06, |
| "loss": 1.1114, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.19912938779290543, |
| "grad_norm": 15.068018913269043, |
| "learning_rate": 6.632716049382717e-06, |
| "loss": 1.0588, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.2037603037880893, |
| "grad_norm": 11.068521499633789, |
| "learning_rate": 6.787037037037037e-06, |
| "loss": 0.9906, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.20839121978327313, |
| "grad_norm": 15.2577486038208, |
| "learning_rate": 6.941358024691358e-06, |
| "loss": 1.0344, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.21302213577845697, |
| "grad_norm": 15.963299751281738, |
| "learning_rate": 7.0956790123456794e-06, |
| "loss": 0.9656, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.21765305177364083, |
| "grad_norm": 12.144145011901855, |
| "learning_rate": 7.25e-06, |
| "loss": 0.9795, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.22228396776882467, |
| "grad_norm": 16.015628814697266, |
| "learning_rate": 7.404320987654322e-06, |
| "loss": 0.8968, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.2269148837640085, |
| "grad_norm": 13.311681747436523, |
| "learning_rate": 7.558641975308643e-06, |
| "loss": 0.9386, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.23154579975919237, |
| "grad_norm": 15.793661117553711, |
| "learning_rate": 7.712962962962964e-06, |
| "loss": 0.9716, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.2361767157543762, |
| "grad_norm": 9.333974838256836, |
| "learning_rate": 7.867283950617284e-06, |
| "loss": 0.9593, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.24080763174956007, |
| "grad_norm": 11.530397415161133, |
| "learning_rate": 8.021604938271605e-06, |
| "loss": 0.9795, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.2454385477447439, |
| "grad_norm": 12.683008193969727, |
| "learning_rate": 8.175925925925925e-06, |
| "loss": 1.0464, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.2500694637399278, |
| "grad_norm": 13.96157455444336, |
| "learning_rate": 8.330246913580248e-06, |
| "loss": 0.9384, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.2547003797351116, |
| "grad_norm": 15.856661796569824, |
| "learning_rate": 8.484567901234568e-06, |
| "loss": 0.9062, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.25933129573029545, |
| "grad_norm": 14.501742362976074, |
| "learning_rate": 8.63888888888889e-06, |
| "loss": 0.8873, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.2639622117254793, |
| "grad_norm": 11.515716552734375, |
| "learning_rate": 8.793209876543211e-06, |
| "loss": 1.0624, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.2685931277206631, |
| "grad_norm": 19.777027130126953, |
| "learning_rate": 8.947530864197532e-06, |
| "loss": 1.0515, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.273224043715847, |
| "grad_norm": 12.990806579589844, |
| "learning_rate": 9.101851851851853e-06, |
| "loss": 0.9574, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.27785495971103086, |
| "grad_norm": 135.59896850585938, |
| "learning_rate": 9.256172839506173e-06, |
| "loss": 0.9481, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.2824858757062147, |
| "grad_norm": 13.494345664978027, |
| "learning_rate": 9.410493827160495e-06, |
| "loss": 1.0147, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.28711679170139853, |
| "grad_norm": 15.373414993286133, |
| "learning_rate": 9.564814814814816e-06, |
| "loss": 0.9794, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.29174770769658237, |
| "grad_norm": 17.2181396484375, |
| "learning_rate": 9.719135802469137e-06, |
| "loss": 0.925, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.2963786236917662, |
| "grad_norm": 15.999231338500977, |
| "learning_rate": 9.873456790123457e-06, |
| "loss": 0.8622, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.3010095396869501, |
| "grad_norm": 12.339484214782715, |
| "learning_rate": 9.996912627354123e-06, |
| "loss": 0.8761, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.30564045568213394, |
| "grad_norm": 12.637123107910156, |
| "learning_rate": 9.979760557099242e-06, |
| "loss": 0.9473, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.3102713716773178, |
| "grad_norm": 11.328486442565918, |
| "learning_rate": 9.962608486844364e-06, |
| "loss": 0.9392, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.3149022876725016, |
| "grad_norm": 16.28173828125, |
| "learning_rate": 9.945456416589483e-06, |
| "loss": 0.871, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.31953320366768545, |
| "grad_norm": 12.468025207519531, |
| "learning_rate": 9.928304346334603e-06, |
| "loss": 0.8408, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.3241641196628693, |
| "grad_norm": 8.507662773132324, |
| "learning_rate": 9.911152276079725e-06, |
| "loss": 0.8601, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.3287950356580532, |
| "grad_norm": 13.56063461303711, |
| "learning_rate": 9.894000205824844e-06, |
| "loss": 0.8586, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.333425951653237, |
| "grad_norm": 13.81701374053955, |
| "learning_rate": 9.876848135569964e-06, |
| "loss": 0.9191, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.33805686764842086, |
| "grad_norm": 14.561142921447754, |
| "learning_rate": 9.859696065315084e-06, |
| "loss": 0.9144, |
| "step": 3650 |
| }, |
| { |
| "epoch": 0.3426877836436047, |
| "grad_norm": 12.986917495727539, |
| "learning_rate": 9.842543995060204e-06, |
| "loss": 0.8209, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.34731869963878853, |
| "grad_norm": 13.493863105773926, |
| "learning_rate": 9.825391924805325e-06, |
| "loss": 0.9236, |
| "step": 3750 |
| }, |
| { |
| "epoch": 0.3519496156339724, |
| "grad_norm": 14.126349449157715, |
| "learning_rate": 9.808239854550445e-06, |
| "loss": 0.8461, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.35658053162915626, |
| "grad_norm": 10.988364219665527, |
| "learning_rate": 9.791087784295566e-06, |
| "loss": 0.8421, |
| "step": 3850 |
| }, |
| { |
| "epoch": 0.3612114476243401, |
| "grad_norm": 11.538872718811035, |
| "learning_rate": 9.773935714040686e-06, |
| "loss": 0.7806, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.36584236361952394, |
| "grad_norm": 19.18027687072754, |
| "learning_rate": 9.756783643785806e-06, |
| "loss": 0.8667, |
| "step": 3950 |
| }, |
| { |
| "epoch": 0.3704732796147078, |
| "grad_norm": 15.992369651794434, |
| "learning_rate": 9.739631573530925e-06, |
| "loss": 0.8681, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.3751041956098916, |
| "grad_norm": 9.232007026672363, |
| "learning_rate": 9.722479503276047e-06, |
| "loss": 0.8877, |
| "step": 4050 |
| }, |
| { |
| "epoch": 0.3797351116050755, |
| "grad_norm": 13.309873580932617, |
| "learning_rate": 9.705327433021166e-06, |
| "loss": 0.8167, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.38436602760025934, |
| "grad_norm": 17.893497467041016, |
| "learning_rate": 9.688175362766286e-06, |
| "loss": 0.8315, |
| "step": 4150 |
| }, |
| { |
| "epoch": 0.3889969435954432, |
| "grad_norm": 16.278261184692383, |
| "learning_rate": 9.671023292511406e-06, |
| "loss": 0.8194, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.393627859590627, |
| "grad_norm": 18.51329803466797, |
| "learning_rate": 9.653871222256527e-06, |
| "loss": 0.8781, |
| "step": 4250 |
| }, |
| { |
| "epoch": 0.39825877558581085, |
| "grad_norm": 13.017523765563965, |
| "learning_rate": 9.636719152001647e-06, |
| "loss": 0.8474, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.40288969158099475, |
| "grad_norm": 13.606826782226562, |
| "learning_rate": 9.619567081746768e-06, |
| "loss": 0.886, |
| "step": 4350 |
| }, |
| { |
| "epoch": 0.4075206075761786, |
| "grad_norm": 15.338058471679688, |
| "learning_rate": 9.602415011491888e-06, |
| "loss": 0.8296, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.4121515235713624, |
| "grad_norm": 15.04211711883545, |
| "learning_rate": 9.585262941237008e-06, |
| "loss": 0.8103, |
| "step": 4450 |
| }, |
| { |
| "epoch": 0.41678243956654626, |
| "grad_norm": 22.55345344543457, |
| "learning_rate": 9.568110870982128e-06, |
| "loss": 0.8835, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.4214133555617301, |
| "grad_norm": 17.12868881225586, |
| "learning_rate": 9.550958800727247e-06, |
| "loss": 0.8298, |
| "step": 4550 |
| }, |
| { |
| "epoch": 0.42604427155691393, |
| "grad_norm": 12.845903396606445, |
| "learning_rate": 9.533806730472369e-06, |
| "loss": 0.8916, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.4306751875520978, |
| "grad_norm": 11.995304107666016, |
| "learning_rate": 9.516654660217489e-06, |
| "loss": 0.8882, |
| "step": 4650 |
| }, |
| { |
| "epoch": 0.43530610354728166, |
| "grad_norm": 15.846125602722168, |
| "learning_rate": 9.49950258996261e-06, |
| "loss": 0.8282, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.4399370195424655, |
| "grad_norm": 10.837459564208984, |
| "learning_rate": 9.48235051970773e-06, |
| "loss": 0.8809, |
| "step": 4750 |
| }, |
| { |
| "epoch": 0.44456793553764934, |
| "grad_norm": 12.318596839904785, |
| "learning_rate": 9.46519844945285e-06, |
| "loss": 0.8353, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.4491988515328332, |
| "grad_norm": 9.558602333068848, |
| "learning_rate": 9.448046379197971e-06, |
| "loss": 0.8337, |
| "step": 4850 |
| }, |
| { |
| "epoch": 0.453829767528017, |
| "grad_norm": 19.20343780517578, |
| "learning_rate": 9.43089430894309e-06, |
| "loss": 0.8147, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.4584606835232009, |
| "grad_norm": 17.662939071655273, |
| "learning_rate": 9.41374223868821e-06, |
| "loss": 0.827, |
| "step": 4950 |
| }, |
| { |
| "epoch": 0.46309159951838474, |
| "grad_norm": 13.364691734313965, |
| "learning_rate": 9.39659016843333e-06, |
| "loss": 0.863, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.4677225155135686, |
| "grad_norm": 13.197148323059082, |
| "learning_rate": 9.37943809817845e-06, |
| "loss": 0.7999, |
| "step": 5050 |
| }, |
| { |
| "epoch": 0.4723534315087524, |
| "grad_norm": 15.571122169494629, |
| "learning_rate": 9.362286027923571e-06, |
| "loss": 0.8459, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.47698434750393626, |
| "grad_norm": 17.007863998413086, |
| "learning_rate": 9.345133957668691e-06, |
| "loss": 0.8159, |
| "step": 5150 |
| }, |
| { |
| "epoch": 0.48161526349912015, |
| "grad_norm": 14.772674560546875, |
| "learning_rate": 9.327981887413812e-06, |
| "loss": 0.8754, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.486246179494304, |
| "grad_norm": 18.637845993041992, |
| "learning_rate": 9.310829817158932e-06, |
| "loss": 0.8453, |
| "step": 5250 |
| }, |
| { |
| "epoch": 0.4908770954894878, |
| "grad_norm": 13.34103775024414, |
| "learning_rate": 9.293677746904052e-06, |
| "loss": 0.8148, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.49550801148467166, |
| "grad_norm": 15.411863327026367, |
| "learning_rate": 9.276525676649172e-06, |
| "loss": 0.7435, |
| "step": 5350 |
| }, |
| { |
| "epoch": 0.5001389274798556, |
| "grad_norm": 11.534259796142578, |
| "learning_rate": 9.259373606394293e-06, |
| "loss": 0.7787, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.5047698434750394, |
| "grad_norm": 10.385927200317383, |
| "learning_rate": 9.242221536139413e-06, |
| "loss": 0.7536, |
| "step": 5450 |
| }, |
| { |
| "epoch": 0.5094007594702232, |
| "grad_norm": 11.001605987548828, |
| "learning_rate": 9.225069465884532e-06, |
| "loss": 0.7829, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.5140316754654071, |
| "grad_norm": 12.940703392028809, |
| "learning_rate": 9.207917395629654e-06, |
| "loss": 0.797, |
| "step": 5550 |
| }, |
| { |
| "epoch": 0.5186625914605909, |
| "grad_norm": 13.790589332580566, |
| "learning_rate": 9.190765325374774e-06, |
| "loss": 0.802, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.5232935074557747, |
| "grad_norm": 14.535255432128906, |
| "learning_rate": 9.173613255119895e-06, |
| "loss": 0.8199, |
| "step": 5650 |
| }, |
| { |
| "epoch": 0.5279244234509586, |
| "grad_norm": 8.654829025268555, |
| "learning_rate": 9.156461184865015e-06, |
| "loss": 0.8698, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.5325553394461424, |
| "grad_norm": 10.871634483337402, |
| "learning_rate": 9.139309114610134e-06, |
| "loss": 0.7877, |
| "step": 5750 |
| }, |
| { |
| "epoch": 0.5371862554413263, |
| "grad_norm": 16.741178512573242, |
| "learning_rate": 9.122157044355254e-06, |
| "loss": 0.7441, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.5418171714365101, |
| "grad_norm": 12.232110023498535, |
| "learning_rate": 9.105004974100374e-06, |
| "loss": 0.7152, |
| "step": 5850 |
| }, |
| { |
| "epoch": 0.546448087431694, |
| "grad_norm": 10.189510345458984, |
| "learning_rate": 9.087852903845494e-06, |
| "loss": 0.7917, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.5510790034268779, |
| "grad_norm": 13.428311347961426, |
| "learning_rate": 9.070700833590615e-06, |
| "loss": 0.7825, |
| "step": 5950 |
| }, |
| { |
| "epoch": 0.5557099194220617, |
| "grad_norm": 7.749415397644043, |
| "learning_rate": 9.053548763335735e-06, |
| "loss": 0.7085, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.5603408354172456, |
| "grad_norm": 12.291046142578125, |
| "learning_rate": 9.036396693080856e-06, |
| "loss": 0.7457, |
| "step": 6050 |
| }, |
| { |
| "epoch": 0.5649717514124294, |
| "grad_norm": 8.931249618530273, |
| "learning_rate": 9.019244622825976e-06, |
| "loss": 0.7621, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.5696026674076132, |
| "grad_norm": 14.065079689025879, |
| "learning_rate": 9.002092552571096e-06, |
| "loss": 0.7981, |
| "step": 6150 |
| }, |
| { |
| "epoch": 0.5742335834027971, |
| "grad_norm": 14.24923038482666, |
| "learning_rate": 8.984940482316217e-06, |
| "loss": 0.7599, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.5788644993979809, |
| "grad_norm": 11.286952018737793, |
| "learning_rate": 8.967788412061337e-06, |
| "loss": 0.764, |
| "step": 6250 |
| }, |
| { |
| "epoch": 0.5834954153931647, |
| "grad_norm": 8.349452018737793, |
| "learning_rate": 8.950636341806457e-06, |
| "loss": 0.7148, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.5881263313883486, |
| "grad_norm": 18.04071617126465, |
| "learning_rate": 8.933484271551576e-06, |
| "loss": 0.7638, |
| "step": 6350 |
| }, |
| { |
| "epoch": 0.5927572473835324, |
| "grad_norm": 12.750236511230469, |
| "learning_rate": 8.916332201296696e-06, |
| "loss": 0.7638, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.5973881633787164, |
| "grad_norm": 10.140752792358398, |
| "learning_rate": 8.899180131041817e-06, |
| "loss": 0.7904, |
| "step": 6450 |
| }, |
| { |
| "epoch": 0.6020190793739002, |
| "grad_norm": 15.348838806152344, |
| "learning_rate": 8.882028060786937e-06, |
| "loss": 0.8024, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.606649995369084, |
| "grad_norm": 12.342340469360352, |
| "learning_rate": 8.864875990532059e-06, |
| "loss": 0.8224, |
| "step": 6550 |
| }, |
| { |
| "epoch": 0.6112809113642679, |
| "grad_norm": 13.80968952178955, |
| "learning_rate": 8.847723920277178e-06, |
| "loss": 0.7567, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.6159118273594517, |
| "grad_norm": 13.79007625579834, |
| "learning_rate": 8.830571850022298e-06, |
| "loss": 0.8385, |
| "step": 6650 |
| }, |
| { |
| "epoch": 0.6205427433546356, |
| "grad_norm": 12.956084251403809, |
| "learning_rate": 8.813419779767418e-06, |
| "loss": 0.7652, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.6251736593498194, |
| "grad_norm": 17.817737579345703, |
| "learning_rate": 8.79626770951254e-06, |
| "loss": 0.7567, |
| "step": 6750 |
| }, |
| { |
| "epoch": 0.6298045753450032, |
| "grad_norm": 12.05063247680664, |
| "learning_rate": 8.779115639257659e-06, |
| "loss": 0.7012, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.6344354913401871, |
| "grad_norm": 7.249739646911621, |
| "learning_rate": 8.761963569002779e-06, |
| "loss": 0.7995, |
| "step": 6850 |
| }, |
| { |
| "epoch": 0.6390664073353709, |
| "grad_norm": 14.997950553894043, |
| "learning_rate": 8.7448114987479e-06, |
| "loss": 0.7259, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.6436973233305547, |
| "grad_norm": 4.541720390319824, |
| "learning_rate": 8.72765942849302e-06, |
| "loss": 0.7801, |
| "step": 6950 |
| }, |
| { |
| "epoch": 0.6483282393257386, |
| "grad_norm": 11.39427661895752, |
| "learning_rate": 8.710507358238141e-06, |
| "loss": 0.7132, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.6529591553209225, |
| "grad_norm": 12.994211196899414, |
| "learning_rate": 8.693355287983261e-06, |
| "loss": 0.7861, |
| "step": 7050 |
| }, |
| { |
| "epoch": 0.6575900713161064, |
| "grad_norm": 13.809148788452148, |
| "learning_rate": 8.67620321772838e-06, |
| "loss": 0.7418, |
| "step": 7100 |
| }, |
| { |
| "epoch": 0.6622209873112902, |
| "grad_norm": 14.191105842590332, |
| "learning_rate": 8.6590511474735e-06, |
| "loss": 0.7191, |
| "step": 7150 |
| }, |
| { |
| "epoch": 0.666851903306474, |
| "grad_norm": 13.280489921569824, |
| "learning_rate": 8.64189907721862e-06, |
| "loss": 0.7834, |
| "step": 7200 |
| }, |
| { |
| "epoch": 0.6714828193016579, |
| "grad_norm": 12.677234649658203, |
| "learning_rate": 8.62474700696374e-06, |
| "loss": 0.752, |
| "step": 7250 |
| }, |
| { |
| "epoch": 0.6761137352968417, |
| "grad_norm": 17.168846130371094, |
| "learning_rate": 8.607594936708861e-06, |
| "loss": 0.7715, |
| "step": 7300 |
| }, |
| { |
| "epoch": 0.6807446512920255, |
| "grad_norm": 16.01129722595215, |
| "learning_rate": 8.590442866453981e-06, |
| "loss": 0.7613, |
| "step": 7350 |
| }, |
| { |
| "epoch": 0.6853755672872094, |
| "grad_norm": 13.499317169189453, |
| "learning_rate": 8.573290796199103e-06, |
| "loss": 0.7688, |
| "step": 7400 |
| }, |
| { |
| "epoch": 0.6900064832823932, |
| "grad_norm": 12.428091049194336, |
| "learning_rate": 8.556138725944222e-06, |
| "loss": 0.6496, |
| "step": 7450 |
| }, |
| { |
| "epoch": 0.6946373992775771, |
| "grad_norm": 13.319454193115234, |
| "learning_rate": 8.538986655689342e-06, |
| "loss": 0.7926, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.6992683152727609, |
| "grad_norm": 16.478046417236328, |
| "learning_rate": 8.521834585434463e-06, |
| "loss": 0.775, |
| "step": 7550 |
| }, |
| { |
| "epoch": 0.7038992312679448, |
| "grad_norm": 17.74137306213379, |
| "learning_rate": 8.504682515179583e-06, |
| "loss": 0.6816, |
| "step": 7600 |
| }, |
| { |
| "epoch": 0.7085301472631287, |
| "grad_norm": 11.713212966918945, |
| "learning_rate": 8.487530444924703e-06, |
| "loss": 0.8058, |
| "step": 7650 |
| }, |
| { |
| "epoch": 0.7131610632583125, |
| "grad_norm": 7.7651872634887695, |
| "learning_rate": 8.470378374669823e-06, |
| "loss": 0.7829, |
| "step": 7700 |
| }, |
| { |
| "epoch": 0.7177919792534964, |
| "grad_norm": 16.14422035217285, |
| "learning_rate": 8.453226304414944e-06, |
| "loss": 0.7515, |
| "step": 7750 |
| }, |
| { |
| "epoch": 0.7224228952486802, |
| "grad_norm": 11.298089981079102, |
| "learning_rate": 8.436074234160064e-06, |
| "loss": 0.7801, |
| "step": 7800 |
| }, |
| { |
| "epoch": 0.727053811243864, |
| "grad_norm": 13.805248260498047, |
| "learning_rate": 8.418922163905185e-06, |
| "loss": 0.7604, |
| "step": 7850 |
| }, |
| { |
| "epoch": 0.7316847272390479, |
| "grad_norm": 14.030163764953613, |
| "learning_rate": 8.401770093650305e-06, |
| "loss": 0.8276, |
| "step": 7900 |
| }, |
| { |
| "epoch": 0.7363156432342317, |
| "grad_norm": 15.09167766571045, |
| "learning_rate": 8.384618023395425e-06, |
| "loss": 0.7044, |
| "step": 7950 |
| }, |
| { |
| "epoch": 0.7409465592294155, |
| "grad_norm": 11.88632583618164, |
| "learning_rate": 8.367465953140544e-06, |
| "loss": 0.6987, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.7455774752245994, |
| "grad_norm": 8.257684707641602, |
| "learning_rate": 8.350313882885664e-06, |
| "loss": 0.7309, |
| "step": 8050 |
| }, |
| { |
| "epoch": 0.7502083912197832, |
| "grad_norm": 16.112545013427734, |
| "learning_rate": 8.333161812630786e-06, |
| "loss": 0.7171, |
| "step": 8100 |
| }, |
| { |
| "epoch": 0.7548393072149672, |
| "grad_norm": 16.633085250854492, |
| "learning_rate": 8.316009742375905e-06, |
| "loss": 0.7689, |
| "step": 8150 |
| }, |
| { |
| "epoch": 0.759470223210151, |
| "grad_norm": 15.874983787536621, |
| "learning_rate": 8.298857672121025e-06, |
| "loss": 0.7185, |
| "step": 8200 |
| }, |
| { |
| "epoch": 0.7641011392053348, |
| "grad_norm": 14.1043119430542, |
| "learning_rate": 8.281705601866146e-06, |
| "loss": 0.7266, |
| "step": 8250 |
| }, |
| { |
| "epoch": 0.7687320552005187, |
| "grad_norm": 9.457536697387695, |
| "learning_rate": 8.264553531611266e-06, |
| "loss": 0.8332, |
| "step": 8300 |
| }, |
| { |
| "epoch": 0.7733629711957025, |
| "grad_norm": 13.245748519897461, |
| "learning_rate": 8.247401461356388e-06, |
| "loss": 0.7343, |
| "step": 8350 |
| }, |
| { |
| "epoch": 0.7779938871908864, |
| "grad_norm": 15.276327133178711, |
| "learning_rate": 8.230249391101507e-06, |
| "loss": 0.7036, |
| "step": 8400 |
| }, |
| { |
| "epoch": 0.7826248031860702, |
| "grad_norm": 7.606396198272705, |
| "learning_rate": 8.213097320846627e-06, |
| "loss": 0.6989, |
| "step": 8450 |
| }, |
| { |
| "epoch": 0.787255719181254, |
| "grad_norm": 11.166284561157227, |
| "learning_rate": 8.195945250591747e-06, |
| "loss": 0.6699, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.7918866351764379, |
| "grad_norm": 13.374491691589355, |
| "learning_rate": 8.178793180336866e-06, |
| "loss": 0.7964, |
| "step": 8550 |
| }, |
| { |
| "epoch": 0.7965175511716217, |
| "grad_norm": 9.178342819213867, |
| "learning_rate": 8.161641110081988e-06, |
| "loss": 0.7728, |
| "step": 8600 |
| }, |
| { |
| "epoch": 0.8011484671668055, |
| "grad_norm": 13.058328628540039, |
| "learning_rate": 8.144489039827108e-06, |
| "loss": 0.778, |
| "step": 8650 |
| }, |
| { |
| "epoch": 0.8057793831619895, |
| "grad_norm": 13.078104019165039, |
| "learning_rate": 8.127336969572227e-06, |
| "loss": 0.8046, |
| "step": 8700 |
| }, |
| { |
| "epoch": 0.8104102991571733, |
| "grad_norm": 8.841740608215332, |
| "learning_rate": 8.110184899317349e-06, |
| "loss": 0.7621, |
| "step": 8750 |
| }, |
| { |
| "epoch": 0.8150412151523572, |
| "grad_norm": 10.117661476135254, |
| "learning_rate": 8.093032829062469e-06, |
| "loss": 0.6871, |
| "step": 8800 |
| }, |
| { |
| "epoch": 0.819672131147541, |
| "grad_norm": 13.581981658935547, |
| "learning_rate": 8.075880758807588e-06, |
| "loss": 0.7598, |
| "step": 8850 |
| }, |
| { |
| "epoch": 0.8243030471427248, |
| "grad_norm": 18.15890884399414, |
| "learning_rate": 8.05872868855271e-06, |
| "loss": 0.7241, |
| "step": 8900 |
| }, |
| { |
| "epoch": 0.8289339631379087, |
| "grad_norm": 10.792105674743652, |
| "learning_rate": 8.04157661829783e-06, |
| "loss": 0.7827, |
| "step": 8950 |
| }, |
| { |
| "epoch": 0.8335648791330925, |
| "grad_norm": 11.7811861038208, |
| "learning_rate": 8.024424548042949e-06, |
| "loss": 0.6763, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.8381957951282764, |
| "grad_norm": 14.736871719360352, |
| "learning_rate": 8.007272477788069e-06, |
| "loss": 0.6663, |
| "step": 9050 |
| }, |
| { |
| "epoch": 0.8428267111234602, |
| "grad_norm": 8.909809112548828, |
| "learning_rate": 7.99012040753319e-06, |
| "loss": 0.7726, |
| "step": 9100 |
| }, |
| { |
| "epoch": 0.847457627118644, |
| "grad_norm": 11.672173500061035, |
| "learning_rate": 7.97296833727831e-06, |
| "loss": 0.7877, |
| "step": 9150 |
| }, |
| { |
| "epoch": 0.8520885431138279, |
| "grad_norm": 17.882083892822266, |
| "learning_rate": 7.955816267023431e-06, |
| "loss": 0.7163, |
| "step": 9200 |
| }, |
| { |
| "epoch": 0.8567194591090118, |
| "grad_norm": 10.789363861083984, |
| "learning_rate": 7.938664196768551e-06, |
| "loss": 0.6821, |
| "step": 9250 |
| }, |
| { |
| "epoch": 0.8613503751041957, |
| "grad_norm": 9.205415725708008, |
| "learning_rate": 7.921512126513671e-06, |
| "loss": 0.7975, |
| "step": 9300 |
| }, |
| { |
| "epoch": 0.8659812910993795, |
| "grad_norm": 10.62431812286377, |
| "learning_rate": 7.90436005625879e-06, |
| "loss": 0.6893, |
| "step": 9350 |
| }, |
| { |
| "epoch": 0.8706122070945633, |
| "grad_norm": 7.258960247039795, |
| "learning_rate": 7.88720798600391e-06, |
| "loss": 0.8111, |
| "step": 9400 |
| }, |
| { |
| "epoch": 0.8752431230897472, |
| "grad_norm": 16.368818283081055, |
| "learning_rate": 7.870055915749032e-06, |
| "loss": 0.7149, |
| "step": 9450 |
| }, |
| { |
| "epoch": 0.879874039084931, |
| "grad_norm": 7.597476482391357, |
| "learning_rate": 7.852903845494151e-06, |
| "loss": 0.6776, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.8845049550801148, |
| "grad_norm": 14.393339157104492, |
| "learning_rate": 7.835751775239271e-06, |
| "loss": 0.6311, |
| "step": 9550 |
| }, |
| { |
| "epoch": 0.8891358710752987, |
| "grad_norm": 11.397239685058594, |
| "learning_rate": 7.818599704984393e-06, |
| "loss": 0.7019, |
| "step": 9600 |
| }, |
| { |
| "epoch": 0.8937667870704825, |
| "grad_norm": 8.419168472290039, |
| "learning_rate": 7.801447634729512e-06, |
| "loss": 0.7242, |
| "step": 9650 |
| }, |
| { |
| "epoch": 0.8983977030656664, |
| "grad_norm": 11.137995719909668, |
| "learning_rate": 7.784295564474634e-06, |
| "loss": 0.6679, |
| "step": 9700 |
| }, |
| { |
| "epoch": 0.9030286190608502, |
| "grad_norm": 12.533097267150879, |
| "learning_rate": 7.767143494219754e-06, |
| "loss": 0.7113, |
| "step": 9750 |
| }, |
| { |
| "epoch": 0.907659535056034, |
| "grad_norm": 8.744980812072754, |
| "learning_rate": 7.749991423964873e-06, |
| "loss": 0.6634, |
| "step": 9800 |
| }, |
| { |
| "epoch": 0.912290451051218, |
| "grad_norm": 12.743056297302246, |
| "learning_rate": 7.732839353709993e-06, |
| "loss": 0.7264, |
| "step": 9850 |
| }, |
| { |
| "epoch": 0.9169213670464018, |
| "grad_norm": 14.499612808227539, |
| "learning_rate": 7.715687283455113e-06, |
| "loss": 0.6815, |
| "step": 9900 |
| }, |
| { |
| "epoch": 0.9215522830415857, |
| "grad_norm": 12.846882820129395, |
| "learning_rate": 7.698535213200234e-06, |
| "loss": 0.7602, |
| "step": 9950 |
| }, |
| { |
| "epoch": 0.9261831990367695, |
| "grad_norm": 10.059849739074707, |
| "learning_rate": 7.681383142945354e-06, |
| "loss": 0.6744, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.9308141150319533, |
| "grad_norm": 13.844738006591797, |
| "learning_rate": 7.664231072690475e-06, |
| "loss": 0.7137, |
| "step": 10050 |
| }, |
| { |
| "epoch": 0.9354450310271372, |
| "grad_norm": 8.811226844787598, |
| "learning_rate": 7.647079002435595e-06, |
| "loss": 0.7499, |
| "step": 10100 |
| }, |
| { |
| "epoch": 0.940075947022321, |
| "grad_norm": 14.390593528747559, |
| "learning_rate": 7.629926932180715e-06, |
| "loss": 0.7108, |
| "step": 10150 |
| }, |
| { |
| "epoch": 0.9447068630175048, |
| "grad_norm": 13.213685035705566, |
| "learning_rate": 7.6127748619258345e-06, |
| "loss": 0.7291, |
| "step": 10200 |
| }, |
| { |
| "epoch": 0.9493377790126887, |
| "grad_norm": 12.737407684326172, |
| "learning_rate": 7.595622791670956e-06, |
| "loss": 0.6997, |
| "step": 10250 |
| }, |
| { |
| "epoch": 0.9539686950078725, |
| "grad_norm": 10.98259449005127, |
| "learning_rate": 7.578470721416076e-06, |
| "loss": 0.7107, |
| "step": 10300 |
| }, |
| { |
| "epoch": 0.9585996110030564, |
| "grad_norm": 12.873851776123047, |
| "learning_rate": 7.561318651161195e-06, |
| "loss": 0.6858, |
| "step": 10350 |
| }, |
| { |
| "epoch": 0.9632305269982403, |
| "grad_norm": 9.490121841430664, |
| "learning_rate": 7.544166580906316e-06, |
| "loss": 0.7393, |
| "step": 10400 |
| }, |
| { |
| "epoch": 0.9678614429934241, |
| "grad_norm": 12.453715324401855, |
| "learning_rate": 7.527014510651436e-06, |
| "loss": 0.7221, |
| "step": 10450 |
| }, |
| { |
| "epoch": 0.972492358988608, |
| "grad_norm": 21.79591941833496, |
| "learning_rate": 7.509862440396557e-06, |
| "loss": 0.7092, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.9771232749837918, |
| "grad_norm": 14.497958183288574, |
| "learning_rate": 7.492710370141677e-06, |
| "loss": 0.7017, |
| "step": 10550 |
| }, |
| { |
| "epoch": 0.9817541909789756, |
| "grad_norm": 16.157657623291016, |
| "learning_rate": 7.475558299886797e-06, |
| "loss": 0.761, |
| "step": 10600 |
| }, |
| { |
| "epoch": 0.9863851069741595, |
| "grad_norm": 9.551992416381836, |
| "learning_rate": 7.458406229631917e-06, |
| "loss": 0.7494, |
| "step": 10650 |
| }, |
| { |
| "epoch": 0.9910160229693433, |
| "grad_norm": 12.097346305847168, |
| "learning_rate": 7.441254159377037e-06, |
| "loss": 0.6412, |
| "step": 10700 |
| }, |
| { |
| "epoch": 0.9956469389645272, |
| "grad_norm": 6.892613887786865, |
| "learning_rate": 7.4241020891221575e-06, |
| "loss": 0.7003, |
| "step": 10750 |
| }, |
| { |
| "epoch": 1.000277854959711, |
| "grad_norm": 10.388261795043945, |
| "learning_rate": 7.406950018867278e-06, |
| "loss": 0.6807, |
| "step": 10800 |
| }, |
| { |
| "epoch": 1.004908770954895, |
| "grad_norm": 7.812435150146484, |
| "learning_rate": 7.389797948612399e-06, |
| "loss": 0.5981, |
| "step": 10850 |
| }, |
| { |
| "epoch": 1.0095396869500788, |
| "grad_norm": 7.479443550109863, |
| "learning_rate": 7.372645878357518e-06, |
| "loss": 0.5875, |
| "step": 10900 |
| }, |
| { |
| "epoch": 1.0141706029452626, |
| "grad_norm": 14.079961776733398, |
| "learning_rate": 7.355493808102638e-06, |
| "loss": 0.6712, |
| "step": 10950 |
| }, |
| { |
| "epoch": 1.0188015189404465, |
| "grad_norm": 13.788300514221191, |
| "learning_rate": 7.338341737847759e-06, |
| "loss": 0.6216, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.0234324349356303, |
| "grad_norm": 10.6414213180542, |
| "learning_rate": 7.321189667592879e-06, |
| "loss": 0.6274, |
| "step": 11050 |
| }, |
| { |
| "epoch": 1.0280633509308141, |
| "grad_norm": 8.648374557495117, |
| "learning_rate": 7.304037597338e-06, |
| "loss": 0.602, |
| "step": 11100 |
| }, |
| { |
| "epoch": 1.032694266925998, |
| "grad_norm": 12.511055946350098, |
| "learning_rate": 7.2868855270831195e-06, |
| "loss": 0.5893, |
| "step": 11150 |
| }, |
| { |
| "epoch": 1.0373251829211818, |
| "grad_norm": 14.385446548461914, |
| "learning_rate": 7.269733456828239e-06, |
| "loss": 0.6545, |
| "step": 11200 |
| }, |
| { |
| "epoch": 1.0419560989163656, |
| "grad_norm": 15.328493118286133, |
| "learning_rate": 7.25258138657336e-06, |
| "loss": 0.6003, |
| "step": 11250 |
| }, |
| { |
| "epoch": 1.0465870149115495, |
| "grad_norm": 13.367624282836914, |
| "learning_rate": 7.23542931631848e-06, |
| "loss": 0.6151, |
| "step": 11300 |
| }, |
| { |
| "epoch": 1.0512179309067333, |
| "grad_norm": 7.708530426025391, |
| "learning_rate": 7.218277246063601e-06, |
| "loss": 0.626, |
| "step": 11350 |
| }, |
| { |
| "epoch": 1.0558488469019172, |
| "grad_norm": 12.538946151733398, |
| "learning_rate": 7.201125175808721e-06, |
| "loss": 0.652, |
| "step": 11400 |
| }, |
| { |
| "epoch": 1.060479762897101, |
| "grad_norm": 12.275097846984863, |
| "learning_rate": 7.1839731055538404e-06, |
| "loss": 0.6398, |
| "step": 11450 |
| }, |
| { |
| "epoch": 1.0651106788922848, |
| "grad_norm": 13.180695533752441, |
| "learning_rate": 7.166821035298961e-06, |
| "loss": 0.6825, |
| "step": 11500 |
| }, |
| { |
| "epoch": 1.0697415948874687, |
| "grad_norm": 15.09621810913086, |
| "learning_rate": 7.149668965044081e-06, |
| "loss": 0.6298, |
| "step": 11550 |
| }, |
| { |
| "epoch": 1.0743725108826525, |
| "grad_norm": 13.7881441116333, |
| "learning_rate": 7.132516894789202e-06, |
| "loss": 0.5414, |
| "step": 11600 |
| }, |
| { |
| "epoch": 1.0790034268778363, |
| "grad_norm": 10.31930160522461, |
| "learning_rate": 7.115364824534322e-06, |
| "loss": 0.6619, |
| "step": 11650 |
| }, |
| { |
| "epoch": 1.0836343428730202, |
| "grad_norm": 10.23030948638916, |
| "learning_rate": 7.0982127542794425e-06, |
| "loss": 0.6333, |
| "step": 11700 |
| }, |
| { |
| "epoch": 1.088265258868204, |
| "grad_norm": 13.737223625183105, |
| "learning_rate": 7.081060684024562e-06, |
| "loss": 0.6896, |
| "step": 11750 |
| }, |
| { |
| "epoch": 1.092896174863388, |
| "grad_norm": 10.776179313659668, |
| "learning_rate": 7.063908613769682e-06, |
| "loss": 0.6393, |
| "step": 11800 |
| }, |
| { |
| "epoch": 1.097527090858572, |
| "grad_norm": 14.287514686584473, |
| "learning_rate": 7.046756543514803e-06, |
| "loss": 0.6185, |
| "step": 11850 |
| }, |
| { |
| "epoch": 1.1021580068537558, |
| "grad_norm": 7.316586017608643, |
| "learning_rate": 7.029604473259923e-06, |
| "loss": 0.5542, |
| "step": 11900 |
| }, |
| { |
| "epoch": 1.1067889228489396, |
| "grad_norm": 10.488934516906738, |
| "learning_rate": 7.012452403005044e-06, |
| "loss": 0.6723, |
| "step": 11950 |
| }, |
| { |
| "epoch": 1.1114198388441234, |
| "grad_norm": 15.451316833496094, |
| "learning_rate": 6.995300332750163e-06, |
| "loss": 0.5714, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.1160507548393073, |
| "grad_norm": 14.99089241027832, |
| "learning_rate": 6.978148262495283e-06, |
| "loss": 0.6006, |
| "step": 12050 |
| }, |
| { |
| "epoch": 1.120681670834491, |
| "grad_norm": 12.084638595581055, |
| "learning_rate": 6.960996192240404e-06, |
| "loss": 0.5881, |
| "step": 12100 |
| }, |
| { |
| "epoch": 1.125312586829675, |
| "grad_norm": 13.337446212768555, |
| "learning_rate": 6.943844121985524e-06, |
| "loss": 0.6301, |
| "step": 12150 |
| }, |
| { |
| "epoch": 1.1299435028248588, |
| "grad_norm": 15.448657989501953, |
| "learning_rate": 6.926692051730645e-06, |
| "loss": 0.6238, |
| "step": 12200 |
| }, |
| { |
| "epoch": 1.1345744188200426, |
| "grad_norm": 8.770259857177734, |
| "learning_rate": 6.909539981475765e-06, |
| "loss": 0.5843, |
| "step": 12250 |
| }, |
| { |
| "epoch": 1.1392053348152265, |
| "grad_norm": 10.406338691711426, |
| "learning_rate": 6.892387911220884e-06, |
| "loss": 0.5978, |
| "step": 12300 |
| }, |
| { |
| "epoch": 1.1438362508104103, |
| "grad_norm": 13.003678321838379, |
| "learning_rate": 6.875235840966005e-06, |
| "loss": 0.6475, |
| "step": 12350 |
| }, |
| { |
| "epoch": 1.1484671668055941, |
| "grad_norm": 13.714174270629883, |
| "learning_rate": 6.8580837707111255e-06, |
| "loss": 0.5861, |
| "step": 12400 |
| }, |
| { |
| "epoch": 1.153098082800778, |
| "grad_norm": 9.3059720993042, |
| "learning_rate": 6.840931700456246e-06, |
| "loss": 0.6743, |
| "step": 12450 |
| }, |
| { |
| "epoch": 1.1577289987959618, |
| "grad_norm": 13.990130424499512, |
| "learning_rate": 6.823779630201366e-06, |
| "loss": 0.6724, |
| "step": 12500 |
| }, |
| { |
| "epoch": 1.1623599147911456, |
| "grad_norm": 13.609382629394531, |
| "learning_rate": 6.8066275599464855e-06, |
| "loss": 0.5501, |
| "step": 12550 |
| }, |
| { |
| "epoch": 1.1669908307863295, |
| "grad_norm": 7.106515884399414, |
| "learning_rate": 6.789475489691606e-06, |
| "loss": 0.5339, |
| "step": 12600 |
| }, |
| { |
| "epoch": 1.1716217467815133, |
| "grad_norm": 17.91675567626953, |
| "learning_rate": 6.7723234194367275e-06, |
| "loss": 0.6094, |
| "step": 12650 |
| }, |
| { |
| "epoch": 1.1762526627766972, |
| "grad_norm": 9.104899406433105, |
| "learning_rate": 6.755171349181847e-06, |
| "loss": 0.6913, |
| "step": 12700 |
| }, |
| { |
| "epoch": 1.180883578771881, |
| "grad_norm": 13.601178169250488, |
| "learning_rate": 6.738019278926967e-06, |
| "loss": 0.6087, |
| "step": 12750 |
| }, |
| { |
| "epoch": 1.1855144947670648, |
| "grad_norm": 12.020017623901367, |
| "learning_rate": 6.7208672086720876e-06, |
| "loss": 0.6051, |
| "step": 12800 |
| }, |
| { |
| "epoch": 1.190145410762249, |
| "grad_norm": 11.219183921813965, |
| "learning_rate": 6.703715138417207e-06, |
| "loss": 0.6096, |
| "step": 12850 |
| }, |
| { |
| "epoch": 1.1947763267574327, |
| "grad_norm": 10.890588760375977, |
| "learning_rate": 6.686563068162327e-06, |
| "loss": 0.5875, |
| "step": 12900 |
| }, |
| { |
| "epoch": 1.1994072427526166, |
| "grad_norm": 9.199064254760742, |
| "learning_rate": 6.6694109979074484e-06, |
| "loss": 0.6276, |
| "step": 12950 |
| }, |
| { |
| "epoch": 1.2040381587478004, |
| "grad_norm": 13.708924293518066, |
| "learning_rate": 6.652258927652568e-06, |
| "loss": 0.608, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.2086690747429842, |
| "grad_norm": 7.073580741882324, |
| "learning_rate": 6.635106857397689e-06, |
| "loss": 0.5901, |
| "step": 13050 |
| }, |
| { |
| "epoch": 1.213299990738168, |
| "grad_norm": 10.898780822753906, |
| "learning_rate": 6.6179547871428085e-06, |
| "loss": 0.554, |
| "step": 13100 |
| }, |
| { |
| "epoch": 1.217930906733352, |
| "grad_norm": 13.31747055053711, |
| "learning_rate": 6.600802716887928e-06, |
| "loss": 0.5498, |
| "step": 13150 |
| }, |
| { |
| "epoch": 1.2225618227285358, |
| "grad_norm": 10.6091890335083, |
| "learning_rate": 6.58365064663305e-06, |
| "loss": 0.6398, |
| "step": 13200 |
| }, |
| { |
| "epoch": 1.2271927387237196, |
| "grad_norm": 11.425141334533691, |
| "learning_rate": 6.566498576378169e-06, |
| "loss": 0.6532, |
| "step": 13250 |
| }, |
| { |
| "epoch": 1.2318236547189034, |
| "grad_norm": 16.974721908569336, |
| "learning_rate": 6.54934650612329e-06, |
| "loss": 0.6373, |
| "step": 13300 |
| }, |
| { |
| "epoch": 1.2364545707140873, |
| "grad_norm": 15.951576232910156, |
| "learning_rate": 6.53219443586841e-06, |
| "loss": 0.5556, |
| "step": 13350 |
| }, |
| { |
| "epoch": 1.241085486709271, |
| "grad_norm": 13.966357231140137, |
| "learning_rate": 6.515042365613529e-06, |
| "loss": 0.6114, |
| "step": 13400 |
| }, |
| { |
| "epoch": 1.245716402704455, |
| "grad_norm": 10.22027587890625, |
| "learning_rate": 6.49789029535865e-06, |
| "loss": 0.5842, |
| "step": 13450 |
| }, |
| { |
| "epoch": 1.2503473186996388, |
| "grad_norm": 14.317681312561035, |
| "learning_rate": 6.4807382251037706e-06, |
| "loss": 0.6064, |
| "step": 13500 |
| }, |
| { |
| "epoch": 1.2549782346948226, |
| "grad_norm": 9.962833404541016, |
| "learning_rate": 6.463586154848891e-06, |
| "loss": 0.5688, |
| "step": 13550 |
| }, |
| { |
| "epoch": 1.2596091506900065, |
| "grad_norm": 5.116006374359131, |
| "learning_rate": 6.446434084594011e-06, |
| "loss": 0.5714, |
| "step": 13600 |
| }, |
| { |
| "epoch": 1.2642400666851903, |
| "grad_norm": 9.99418830871582, |
| "learning_rate": 6.4292820143391314e-06, |
| "loss": 0.6329, |
| "step": 13650 |
| }, |
| { |
| "epoch": 1.2688709826803741, |
| "grad_norm": 12.337767601013184, |
| "learning_rate": 6.412129944084251e-06, |
| "loss": 0.5907, |
| "step": 13700 |
| }, |
| { |
| "epoch": 1.273501898675558, |
| "grad_norm": 7.793351173400879, |
| "learning_rate": 6.394977873829373e-06, |
| "loss": 0.5418, |
| "step": 13750 |
| }, |
| { |
| "epoch": 1.2781328146707418, |
| "grad_norm": 6.3952436447143555, |
| "learning_rate": 6.377825803574492e-06, |
| "loss": 0.6304, |
| "step": 13800 |
| }, |
| { |
| "epoch": 1.2827637306659256, |
| "grad_norm": 10.685275077819824, |
| "learning_rate": 6.360673733319612e-06, |
| "loss": 0.5713, |
| "step": 13850 |
| }, |
| { |
| "epoch": 1.2873946466611095, |
| "grad_norm": 12.462202072143555, |
| "learning_rate": 6.343521663064733e-06, |
| "loss": 0.64, |
| "step": 13900 |
| }, |
| { |
| "epoch": 1.2920255626562933, |
| "grad_norm": 11.661152839660645, |
| "learning_rate": 6.326369592809852e-06, |
| "loss": 0.7054, |
| "step": 13950 |
| }, |
| { |
| "epoch": 1.2966564786514772, |
| "grad_norm": 11.020445823669434, |
| "learning_rate": 6.309217522554974e-06, |
| "loss": 0.6831, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.301287394646661, |
| "grad_norm": 12.878907203674316, |
| "learning_rate": 6.2920654523000935e-06, |
| "loss": 0.5805, |
| "step": 14050 |
| }, |
| { |
| "epoch": 1.3059183106418448, |
| "grad_norm": 7.879404544830322, |
| "learning_rate": 6.274913382045213e-06, |
| "loss": 0.5727, |
| "step": 14100 |
| }, |
| { |
| "epoch": 1.3105492266370289, |
| "grad_norm": 13.666654586791992, |
| "learning_rate": 6.257761311790334e-06, |
| "loss": 0.5393, |
| "step": 14150 |
| }, |
| { |
| "epoch": 1.3151801426322127, |
| "grad_norm": 13.355966567993164, |
| "learning_rate": 6.2406092415354535e-06, |
| "loss": 0.5949, |
| "step": 14200 |
| }, |
| { |
| "epoch": 1.3198110586273966, |
| "grad_norm": 9.979362487792969, |
| "learning_rate": 6.223457171280573e-06, |
| "loss": 0.5794, |
| "step": 14250 |
| }, |
| { |
| "epoch": 1.3244419746225804, |
| "grad_norm": 14.988487243652344, |
| "learning_rate": 6.206305101025695e-06, |
| "loss": 0.61, |
| "step": 14300 |
| }, |
| { |
| "epoch": 1.3290728906177642, |
| "grad_norm": 13.34450626373291, |
| "learning_rate": 6.1891530307708144e-06, |
| "loss": 0.6704, |
| "step": 14350 |
| }, |
| { |
| "epoch": 1.333703806612948, |
| "grad_norm": 12.021376609802246, |
| "learning_rate": 6.172000960515935e-06, |
| "loss": 0.5829, |
| "step": 14400 |
| }, |
| { |
| "epoch": 1.338334722608132, |
| "grad_norm": 11.892980575561523, |
| "learning_rate": 6.154848890261055e-06, |
| "loss": 0.5603, |
| "step": 14450 |
| }, |
| { |
| "epoch": 1.3429656386033157, |
| "grad_norm": 15.183904647827148, |
| "learning_rate": 6.1376968200061745e-06, |
| "loss": 0.6163, |
| "step": 14500 |
| }, |
| { |
| "epoch": 1.3475965545984996, |
| "grad_norm": 14.723098754882812, |
| "learning_rate": 6.120544749751296e-06, |
| "loss": 0.7141, |
| "step": 14550 |
| }, |
| { |
| "epoch": 1.3522274705936834, |
| "grad_norm": 7.206448554992676, |
| "learning_rate": 6.103392679496416e-06, |
| "loss": 0.5413, |
| "step": 14600 |
| }, |
| { |
| "epoch": 1.3568583865888673, |
| "grad_norm": 14.650779724121094, |
| "learning_rate": 6.086240609241536e-06, |
| "loss": 0.5972, |
| "step": 14650 |
| }, |
| { |
| "epoch": 1.361489302584051, |
| "grad_norm": 11.455196380615234, |
| "learning_rate": 6.069088538986656e-06, |
| "loss": 0.6104, |
| "step": 14700 |
| }, |
| { |
| "epoch": 1.366120218579235, |
| "grad_norm": 10.497644424438477, |
| "learning_rate": 6.0519364687317765e-06, |
| "loss": 0.6147, |
| "step": 14750 |
| }, |
| { |
| "epoch": 1.3707511345744188, |
| "grad_norm": 15.974631309509277, |
| "learning_rate": 6.034784398476896e-06, |
| "loss": 0.6319, |
| "step": 14800 |
| }, |
| { |
| "epoch": 1.3753820505696026, |
| "grad_norm": 13.41238021850586, |
| "learning_rate": 6.017632328222018e-06, |
| "loss": 0.6277, |
| "step": 14850 |
| }, |
| { |
| "epoch": 1.3800129665647864, |
| "grad_norm": 11.693626403808594, |
| "learning_rate": 6.000480257967137e-06, |
| "loss": 0.5986, |
| "step": 14900 |
| }, |
| { |
| "epoch": 1.3846438825599703, |
| "grad_norm": 14.90977668762207, |
| "learning_rate": 5.983328187712257e-06, |
| "loss": 0.5409, |
| "step": 14950 |
| }, |
| { |
| "epoch": 1.3892747985551543, |
| "grad_norm": 15.576924324035645, |
| "learning_rate": 5.966176117457378e-06, |
| "loss": 0.6396, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.3939057145503382, |
| "grad_norm": 12.302413940429688, |
| "learning_rate": 5.9490240472024974e-06, |
| "loss": 0.599, |
| "step": 15050 |
| }, |
| { |
| "epoch": 1.398536630545522, |
| "grad_norm": 11.971318244934082, |
| "learning_rate": 5.931871976947619e-06, |
| "loss": 0.6512, |
| "step": 15100 |
| }, |
| { |
| "epoch": 1.4031675465407059, |
| "grad_norm": 9.926959037780762, |
| "learning_rate": 5.914719906692739e-06, |
| "loss": 0.5849, |
| "step": 15150 |
| }, |
| { |
| "epoch": 1.4077984625358897, |
| "grad_norm": 4.981182098388672, |
| "learning_rate": 5.897567836437858e-06, |
| "loss": 0.582, |
| "step": 15200 |
| }, |
| { |
| "epoch": 1.4124293785310735, |
| "grad_norm": 13.111713409423828, |
| "learning_rate": 5.880415766182979e-06, |
| "loss": 0.6674, |
| "step": 15250 |
| }, |
| { |
| "epoch": 1.4170602945262574, |
| "grad_norm": 14.456388473510742, |
| "learning_rate": 5.863263695928099e-06, |
| "loss": 0.6413, |
| "step": 15300 |
| }, |
| { |
| "epoch": 1.4216912105214412, |
| "grad_norm": 12.475411415100098, |
| "learning_rate": 5.84611162567322e-06, |
| "loss": 0.6454, |
| "step": 15350 |
| }, |
| { |
| "epoch": 1.426322126516625, |
| "grad_norm": 14.428534507751465, |
| "learning_rate": 5.82895955541834e-06, |
| "loss": 0.5802, |
| "step": 15400 |
| }, |
| { |
| "epoch": 1.4309530425118089, |
| "grad_norm": 10.774445533752441, |
| "learning_rate": 5.8118074851634595e-06, |
| "loss": 0.6026, |
| "step": 15450 |
| }, |
| { |
| "epoch": 1.4355839585069927, |
| "grad_norm": 7.127171039581299, |
| "learning_rate": 5.79465541490858e-06, |
| "loss": 0.5861, |
| "step": 15500 |
| }, |
| { |
| "epoch": 1.4402148745021766, |
| "grad_norm": 20.572057723999023, |
| "learning_rate": 5.7775033446537e-06, |
| "loss": 0.6995, |
| "step": 15550 |
| }, |
| { |
| "epoch": 1.4448457904973604, |
| "grad_norm": 9.042718887329102, |
| "learning_rate": 5.7603512743988195e-06, |
| "loss": 0.602, |
| "step": 15600 |
| }, |
| { |
| "epoch": 1.4494767064925442, |
| "grad_norm": 10.351805686950684, |
| "learning_rate": 5.743199204143941e-06, |
| "loss": 0.5973, |
| "step": 15650 |
| }, |
| { |
| "epoch": 1.454107622487728, |
| "grad_norm": 20.35566520690918, |
| "learning_rate": 5.726047133889061e-06, |
| "loss": 0.5642, |
| "step": 15700 |
| }, |
| { |
| "epoch": 1.458738538482912, |
| "grad_norm": 12.212347984313965, |
| "learning_rate": 5.708895063634181e-06, |
| "loss": 0.5624, |
| "step": 15750 |
| }, |
| { |
| "epoch": 1.4633694544780957, |
| "grad_norm": 7.124228477478027, |
| "learning_rate": 5.691742993379301e-06, |
| "loss": 0.5613, |
| "step": 15800 |
| }, |
| { |
| "epoch": 1.4680003704732796, |
| "grad_norm": 18.672344207763672, |
| "learning_rate": 5.6745909231244216e-06, |
| "loss": 0.5946, |
| "step": 15850 |
| }, |
| { |
| "epoch": 1.4726312864684634, |
| "grad_norm": 15.380640029907227, |
| "learning_rate": 5.657438852869542e-06, |
| "loss": 0.6391, |
| "step": 15900 |
| }, |
| { |
| "epoch": 1.4772622024636473, |
| "grad_norm": 13.782407760620117, |
| "learning_rate": 5.640286782614663e-06, |
| "loss": 0.5824, |
| "step": 15950 |
| }, |
| { |
| "epoch": 1.481893118458831, |
| "grad_norm": 10.316747665405273, |
| "learning_rate": 5.6231347123597825e-06, |
| "loss": 0.5761, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.486524034454015, |
| "grad_norm": 14.089554786682129, |
| "learning_rate": 5.605982642104902e-06, |
| "loss": 0.6171, |
| "step": 16050 |
| }, |
| { |
| "epoch": 1.4911549504491988, |
| "grad_norm": 11.800119400024414, |
| "learning_rate": 5.588830571850023e-06, |
| "loss": 0.6445, |
| "step": 16100 |
| }, |
| { |
| "epoch": 1.4957858664443826, |
| "grad_norm": 13.505830764770508, |
| "learning_rate": 5.5716785015951425e-06, |
| "loss": 0.5833, |
| "step": 16150 |
| }, |
| { |
| "epoch": 1.5004167824395664, |
| "grad_norm": 11.643794059753418, |
| "learning_rate": 5.554526431340264e-06, |
| "loss": 0.5479, |
| "step": 16200 |
| }, |
| { |
| "epoch": 1.5050476984347503, |
| "grad_norm": 16.229856491088867, |
| "learning_rate": 5.537374361085384e-06, |
| "loss": 0.603, |
| "step": 16250 |
| }, |
| { |
| "epoch": 1.5096786144299341, |
| "grad_norm": 23.821744918823242, |
| "learning_rate": 5.520222290830503e-06, |
| "loss": 0.5961, |
| "step": 16300 |
| }, |
| { |
| "epoch": 1.514309530425118, |
| "grad_norm": 16.5296630859375, |
| "learning_rate": 5.503070220575624e-06, |
| "loss": 0.5738, |
| "step": 16350 |
| }, |
| { |
| "epoch": 1.5189404464203018, |
| "grad_norm": 14.164803504943848, |
| "learning_rate": 5.485918150320744e-06, |
| "loss": 0.5495, |
| "step": 16400 |
| }, |
| { |
| "epoch": 1.5235713624154856, |
| "grad_norm": 10.112049102783203, |
| "learning_rate": 5.468766080065865e-06, |
| "loss": 0.5656, |
| "step": 16450 |
| }, |
| { |
| "epoch": 1.5282022784106695, |
| "grad_norm": 11.996503829956055, |
| "learning_rate": 5.451614009810985e-06, |
| "loss": 0.6021, |
| "step": 16500 |
| }, |
| { |
| "epoch": 1.5328331944058535, |
| "grad_norm": 7.8670525550842285, |
| "learning_rate": 5.4344619395561046e-06, |
| "loss": 0.6592, |
| "step": 16550 |
| }, |
| { |
| "epoch": 1.5374641104010374, |
| "grad_norm": 14.625728607177734, |
| "learning_rate": 5.417309869301225e-06, |
| "loss": 0.6173, |
| "step": 16600 |
| }, |
| { |
| "epoch": 1.5420950263962212, |
| "grad_norm": 14.137253761291504, |
| "learning_rate": 5.400157799046345e-06, |
| "loss": 0.5709, |
| "step": 16650 |
| }, |
| { |
| "epoch": 1.546725942391405, |
| "grad_norm": 9.287371635437012, |
| "learning_rate": 5.383005728791466e-06, |
| "loss": 0.6174, |
| "step": 16700 |
| }, |
| { |
| "epoch": 1.5513568583865889, |
| "grad_norm": 12.439033508300781, |
| "learning_rate": 5.365853658536586e-06, |
| "loss": 0.5798, |
| "step": 16750 |
| }, |
| { |
| "epoch": 1.5559877743817727, |
| "grad_norm": 7.709993362426758, |
| "learning_rate": 5.348701588281706e-06, |
| "loss": 0.571, |
| "step": 16800 |
| }, |
| { |
| "epoch": 1.5606186903769566, |
| "grad_norm": 8.796138763427734, |
| "learning_rate": 5.331549518026826e-06, |
| "loss": 0.5744, |
| "step": 16850 |
| }, |
| { |
| "epoch": 1.5652496063721404, |
| "grad_norm": 7.53299617767334, |
| "learning_rate": 5.314397447771946e-06, |
| "loss": 0.6213, |
| "step": 16900 |
| }, |
| { |
| "epoch": 1.5698805223673242, |
| "grad_norm": 13.07565689086914, |
| "learning_rate": 5.297245377517067e-06, |
| "loss": 0.586, |
| "step": 16950 |
| }, |
| { |
| "epoch": 1.574511438362508, |
| "grad_norm": 8.85544490814209, |
| "learning_rate": 5.280093307262187e-06, |
| "loss": 0.5868, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.579142354357692, |
| "grad_norm": 11.888114929199219, |
| "learning_rate": 5.262941237007308e-06, |
| "loss": 0.6038, |
| "step": 17050 |
| }, |
| { |
| "epoch": 1.583773270352876, |
| "grad_norm": 15.81039047241211, |
| "learning_rate": 5.2457891667524275e-06, |
| "loss": 0.5684, |
| "step": 17100 |
| }, |
| { |
| "epoch": 1.5884041863480598, |
| "grad_norm": 11.658292770385742, |
| "learning_rate": 5.228637096497547e-06, |
| "loss": 0.63, |
| "step": 17150 |
| }, |
| { |
| "epoch": 1.5930351023432436, |
| "grad_norm": 12.549324989318848, |
| "learning_rate": 5.211485026242668e-06, |
| "loss": 0.5832, |
| "step": 17200 |
| }, |
| { |
| "epoch": 1.5976660183384275, |
| "grad_norm": 13.756429672241211, |
| "learning_rate": 5.194332955987788e-06, |
| "loss": 0.566, |
| "step": 17250 |
| }, |
| { |
| "epoch": 1.6022969343336113, |
| "grad_norm": 13.937918663024902, |
| "learning_rate": 5.177180885732909e-06, |
| "loss": 0.5913, |
| "step": 17300 |
| }, |
| { |
| "epoch": 1.6069278503287951, |
| "grad_norm": 15.999268531799316, |
| "learning_rate": 5.160028815478029e-06, |
| "loss": 0.6455, |
| "step": 17350 |
| }, |
| { |
| "epoch": 1.611558766323979, |
| "grad_norm": 11.241209983825684, |
| "learning_rate": 5.1428767452231485e-06, |
| "loss": 0.5808, |
| "step": 17400 |
| }, |
| { |
| "epoch": 1.6161896823191628, |
| "grad_norm": 10.14393424987793, |
| "learning_rate": 5.125724674968269e-06, |
| "loss": 0.54, |
| "step": 17450 |
| }, |
| { |
| "epoch": 1.6208205983143467, |
| "grad_norm": 10.87530517578125, |
| "learning_rate": 5.108572604713389e-06, |
| "loss": 0.6425, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.6254515143095305, |
| "grad_norm": 14.635629653930664, |
| "learning_rate": 5.09142053445851e-06, |
| "loss": 0.5994, |
| "step": 17550 |
| }, |
| { |
| "epoch": 1.6300824303047143, |
| "grad_norm": 17.629318237304688, |
| "learning_rate": 5.07426846420363e-06, |
| "loss": 0.5827, |
| "step": 17600 |
| }, |
| { |
| "epoch": 1.6347133462998982, |
| "grad_norm": 14.779356002807617, |
| "learning_rate": 5.05711639394875e-06, |
| "loss": 0.57, |
| "step": 17650 |
| }, |
| { |
| "epoch": 1.639344262295082, |
| "grad_norm": 13.21997356414795, |
| "learning_rate": 5.03996432369387e-06, |
| "loss": 0.555, |
| "step": 17700 |
| }, |
| { |
| "epoch": 1.6439751782902658, |
| "grad_norm": 10.608312606811523, |
| "learning_rate": 5.02281225343899e-06, |
| "loss": 0.6238, |
| "step": 17750 |
| }, |
| { |
| "epoch": 1.6486060942854497, |
| "grad_norm": 10.145637512207031, |
| "learning_rate": 5.005660183184111e-06, |
| "loss": 0.6148, |
| "step": 17800 |
| }, |
| { |
| "epoch": 1.6532370102806335, |
| "grad_norm": 13.766275405883789, |
| "learning_rate": 4.988508112929231e-06, |
| "loss": 0.6019, |
| "step": 17850 |
| }, |
| { |
| "epoch": 1.6578679262758174, |
| "grad_norm": 12.229156494140625, |
| "learning_rate": 4.971356042674351e-06, |
| "loss": 0.6072, |
| "step": 17900 |
| }, |
| { |
| "epoch": 1.6624988422710012, |
| "grad_norm": 13.928374290466309, |
| "learning_rate": 4.954203972419471e-06, |
| "loss": 0.6248, |
| "step": 17950 |
| }, |
| { |
| "epoch": 1.667129758266185, |
| "grad_norm": 11.471216201782227, |
| "learning_rate": 4.937051902164592e-06, |
| "loss": 0.6054, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.6717606742613689, |
| "grad_norm": 15.625840187072754, |
| "learning_rate": 4.919899831909712e-06, |
| "loss": 0.5306, |
| "step": 18050 |
| }, |
| { |
| "epoch": 1.6763915902565527, |
| "grad_norm": 13.443790435791016, |
| "learning_rate": 4.902747761654832e-06, |
| "loss": 0.574, |
| "step": 18100 |
| }, |
| { |
| "epoch": 1.6810225062517365, |
| "grad_norm": 9.181079864501953, |
| "learning_rate": 4.885595691399953e-06, |
| "loss": 0.5879, |
| "step": 18150 |
| }, |
| { |
| "epoch": 1.6856534222469204, |
| "grad_norm": 11.0618314743042, |
| "learning_rate": 4.868443621145073e-06, |
| "loss": 0.5564, |
| "step": 18200 |
| }, |
| { |
| "epoch": 1.6902843382421042, |
| "grad_norm": 11.787243843078613, |
| "learning_rate": 4.851291550890193e-06, |
| "loss": 0.5756, |
| "step": 18250 |
| }, |
| { |
| "epoch": 1.694915254237288, |
| "grad_norm": 14.000703811645508, |
| "learning_rate": 4.834139480635313e-06, |
| "loss": 0.5347, |
| "step": 18300 |
| }, |
| { |
| "epoch": 1.699546170232472, |
| "grad_norm": 17.239091873168945, |
| "learning_rate": 4.8169874103804335e-06, |
| "loss": 0.5733, |
| "step": 18350 |
| }, |
| { |
| "epoch": 1.7041770862276557, |
| "grad_norm": 9.486557006835938, |
| "learning_rate": 4.799835340125554e-06, |
| "loss": 0.6062, |
| "step": 18400 |
| }, |
| { |
| "epoch": 1.7088080022228396, |
| "grad_norm": 8.935564994812012, |
| "learning_rate": 4.782683269870674e-06, |
| "loss": 0.5582, |
| "step": 18450 |
| }, |
| { |
| "epoch": 1.7134389182180234, |
| "grad_norm": 12.081872940063477, |
| "learning_rate": 4.7655311996157935e-06, |
| "loss": 0.5848, |
| "step": 18500 |
| }, |
| { |
| "epoch": 1.7180698342132072, |
| "grad_norm": 11.821228981018066, |
| "learning_rate": 4.748379129360914e-06, |
| "loss": 0.537, |
| "step": 18550 |
| }, |
| { |
| "epoch": 1.722700750208391, |
| "grad_norm": 17.22193145751953, |
| "learning_rate": 4.731227059106035e-06, |
| "loss": 0.5867, |
| "step": 18600 |
| }, |
| { |
| "epoch": 1.727331666203575, |
| "grad_norm": 14.369434356689453, |
| "learning_rate": 4.714074988851154e-06, |
| "loss": 0.5729, |
| "step": 18650 |
| }, |
| { |
| "epoch": 1.7319625821987588, |
| "grad_norm": 8.231170654296875, |
| "learning_rate": 4.696922918596275e-06, |
| "loss": 0.6438, |
| "step": 18700 |
| }, |
| { |
| "epoch": 1.7365934981939426, |
| "grad_norm": 9.67603874206543, |
| "learning_rate": 4.679770848341395e-06, |
| "loss": 0.5475, |
| "step": 18750 |
| }, |
| { |
| "epoch": 1.7412244141891267, |
| "grad_norm": 13.235024452209473, |
| "learning_rate": 4.662618778086515e-06, |
| "loss": 0.6442, |
| "step": 18800 |
| }, |
| { |
| "epoch": 1.7458553301843105, |
| "grad_norm": 9.636026382446289, |
| "learning_rate": 4.645466707831636e-06, |
| "loss": 0.5909, |
| "step": 18850 |
| }, |
| { |
| "epoch": 1.7504862461794943, |
| "grad_norm": 14.982551574707031, |
| "learning_rate": 4.628314637576756e-06, |
| "loss": 0.5283, |
| "step": 18900 |
| }, |
| { |
| "epoch": 1.7551171621746782, |
| "grad_norm": 13.695237159729004, |
| "learning_rate": 4.611162567321876e-06, |
| "loss": 0.5789, |
| "step": 18950 |
| }, |
| { |
| "epoch": 1.759748078169862, |
| "grad_norm": 16.833280563354492, |
| "learning_rate": 4.594010497066996e-06, |
| "loss": 0.604, |
| "step": 19000 |
| }, |
| { |
| "epoch": 1.7643789941650458, |
| "grad_norm": 8.901232719421387, |
| "learning_rate": 4.5768584268121165e-06, |
| "loss": 0.5661, |
| "step": 19050 |
| }, |
| { |
| "epoch": 1.7690099101602297, |
| "grad_norm": 8.41875171661377, |
| "learning_rate": 4.559706356557237e-06, |
| "loss": 0.5905, |
| "step": 19100 |
| }, |
| { |
| "epoch": 1.7736408261554135, |
| "grad_norm": 12.987443923950195, |
| "learning_rate": 4.542554286302357e-06, |
| "loss": 0.5465, |
| "step": 19150 |
| }, |
| { |
| "epoch": 1.7782717421505974, |
| "grad_norm": 7.876839637756348, |
| "learning_rate": 4.525402216047477e-06, |
| "loss": 0.5171, |
| "step": 19200 |
| }, |
| { |
| "epoch": 1.7829026581457812, |
| "grad_norm": 7.976441860198975, |
| "learning_rate": 4.508250145792598e-06, |
| "loss": 0.5637, |
| "step": 19250 |
| }, |
| { |
| "epoch": 1.787533574140965, |
| "grad_norm": 11.731218338012695, |
| "learning_rate": 4.491098075537718e-06, |
| "loss": 0.6353, |
| "step": 19300 |
| }, |
| { |
| "epoch": 1.792164490136149, |
| "grad_norm": 12.95614242553711, |
| "learning_rate": 4.473946005282838e-06, |
| "loss": 0.5705, |
| "step": 19350 |
| }, |
| { |
| "epoch": 1.796795406131333, |
| "grad_norm": 12.79031753540039, |
| "learning_rate": 4.456793935027958e-06, |
| "loss": 0.6166, |
| "step": 19400 |
| }, |
| { |
| "epoch": 1.8014263221265168, |
| "grad_norm": 5.699042797088623, |
| "learning_rate": 4.4396418647730786e-06, |
| "loss": 0.5323, |
| "step": 19450 |
| }, |
| { |
| "epoch": 1.8060572381217006, |
| "grad_norm": 12.937334060668945, |
| "learning_rate": 4.422489794518199e-06, |
| "loss": 0.5709, |
| "step": 19500 |
| }, |
| { |
| "epoch": 1.8106881541168844, |
| "grad_norm": 14.37783145904541, |
| "learning_rate": 4.405337724263319e-06, |
| "loss": 0.5967, |
| "step": 19550 |
| }, |
| { |
| "epoch": 1.8153190701120683, |
| "grad_norm": 11.934647560119629, |
| "learning_rate": 4.3881856540084394e-06, |
| "loss": 0.5723, |
| "step": 19600 |
| }, |
| { |
| "epoch": 1.8199499861072521, |
| "grad_norm": 10.981630325317383, |
| "learning_rate": 4.371033583753559e-06, |
| "loss": 0.5652, |
| "step": 19650 |
| }, |
| { |
| "epoch": 1.824580902102436, |
| "grad_norm": 17.21076202392578, |
| "learning_rate": 4.35388151349868e-06, |
| "loss": 0.5898, |
| "step": 19700 |
| }, |
| { |
| "epoch": 1.8292118180976198, |
| "grad_norm": 13.629597663879395, |
| "learning_rate": 4.3367294432438e-06, |
| "loss": 0.6095, |
| "step": 19750 |
| }, |
| { |
| "epoch": 1.8338427340928036, |
| "grad_norm": 8.095635414123535, |
| "learning_rate": 4.31957737298892e-06, |
| "loss": 0.5825, |
| "step": 19800 |
| }, |
| { |
| "epoch": 1.8384736500879875, |
| "grad_norm": 8.601015090942383, |
| "learning_rate": 4.30242530273404e-06, |
| "loss": 0.587, |
| "step": 19850 |
| }, |
| { |
| "epoch": 1.8431045660831713, |
| "grad_norm": 13.996216773986816, |
| "learning_rate": 4.28527323247916e-06, |
| "loss": 0.5412, |
| "step": 19900 |
| }, |
| { |
| "epoch": 1.8477354820783551, |
| "grad_norm": 7.266920566558838, |
| "learning_rate": 4.268121162224281e-06, |
| "loss": 0.5759, |
| "step": 19950 |
| }, |
| { |
| "epoch": 1.852366398073539, |
| "grad_norm": 12.852421760559082, |
| "learning_rate": 4.2509690919694015e-06, |
| "loss": 0.559, |
| "step": 20000 |
| } |
| ], |
| "logging_steps": 50, |
| "max_steps": 32391, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 5000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
