| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.0, |
| "eval_steps": 500, |
| "global_step": 3624, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0027597626604112047, |
| "grad_norm": 9.687899712924036, |
| "learning_rate": 2.4793388429752067e-07, |
| "loss": 1.2489, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.005519525320822409, |
| "grad_norm": 7.993583387794843, |
| "learning_rate": 5.234159779614326e-07, |
| "loss": 1.23, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.008279287981233613, |
| "grad_norm": 4.419911118410825, |
| "learning_rate": 7.988980716253444e-07, |
| "loss": 1.134, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.011039050641644819, |
| "grad_norm": 3.0732247545437965, |
| "learning_rate": 1.0743801652892562e-06, |
| "loss": 0.993, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.013798813302056023, |
| "grad_norm": 2.005789581836949, |
| "learning_rate": 1.3498622589531682e-06, |
| "loss": 0.8782, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.016558575962467226, |
| "grad_norm": 1.3028620164242184, |
| "learning_rate": 1.62534435261708e-06, |
| "loss": 0.8004, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.019318338622878434, |
| "grad_norm": 1.2871645472343145, |
| "learning_rate": 1.900826446280992e-06, |
| "loss": 0.7618, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.022078101283289638, |
| "grad_norm": 2.0977280137118854, |
| "learning_rate": 2.1763085399449038e-06, |
| "loss": 0.7363, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.02483786394370084, |
| "grad_norm": 1.235679470482907, |
| "learning_rate": 2.4517906336088157e-06, |
| "loss": 0.7218, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.027597626604112045, |
| "grad_norm": 1.2008847267657137, |
| "learning_rate": 2.7272727272727272e-06, |
| "loss": 0.7039, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.030357389264523253, |
| "grad_norm": 1.2865692124934534, |
| "learning_rate": 3.002754820936639e-06, |
| "loss": 0.7016, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.03311715192493445, |
| "grad_norm": 1.225519692241417, |
| "learning_rate": 3.278236914600551e-06, |
| "loss": 0.6911, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.035876914585345664, |
| "grad_norm": 1.1074744380119828, |
| "learning_rate": 3.553719008264463e-06, |
| "loss": 0.6837, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.03863667724575687, |
| "grad_norm": 1.2255853901709166, |
| "learning_rate": 3.8292011019283746e-06, |
| "loss": 0.6733, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.04139643990616807, |
| "grad_norm": 1.2160790653325682, |
| "learning_rate": 4.104683195592287e-06, |
| "loss": 0.6663, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.044156202566579275, |
| "grad_norm": 1.2160126107257938, |
| "learning_rate": 4.3801652892561984e-06, |
| "loss": 0.6672, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.04691596522699048, |
| "grad_norm": 1.3030345619512333, |
| "learning_rate": 4.655647382920111e-06, |
| "loss": 0.6604, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.04967572788740168, |
| "grad_norm": 1.1675762652939756, |
| "learning_rate": 4.931129476584022e-06, |
| "loss": 0.654, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.05243549054781289, |
| "grad_norm": 1.3421701645033057, |
| "learning_rate": 5.206611570247935e-06, |
| "loss": 0.653, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.05519525320822409, |
| "grad_norm": 1.1987056035659454, |
| "learning_rate": 5.482093663911846e-06, |
| "loss": 0.6508, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.057955015868635294, |
| "grad_norm": 1.3869845490974586, |
| "learning_rate": 5.7575757575757586e-06, |
| "loss": 0.6498, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.060714778529046505, |
| "grad_norm": 1.116588718762935, |
| "learning_rate": 6.03305785123967e-06, |
| "loss": 0.6453, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.06347454118945771, |
| "grad_norm": 1.0960389471025322, |
| "learning_rate": 6.3085399449035824e-06, |
| "loss": 0.6414, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.0662343038498689, |
| "grad_norm": 1.2133969264114732, |
| "learning_rate": 6.584022038567494e-06, |
| "loss": 0.6345, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.06899406651028012, |
| "grad_norm": 1.2700545101041685, |
| "learning_rate": 6.859504132231406e-06, |
| "loss": 0.6386, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.07175382917069133, |
| "grad_norm": 1.2174406212987252, |
| "learning_rate": 7.134986225895317e-06, |
| "loss": 0.6403, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.07451359183110252, |
| "grad_norm": 1.1500811137175568, |
| "learning_rate": 7.410468319559229e-06, |
| "loss": 0.6361, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.07727335449151373, |
| "grad_norm": 1.2075741428461082, |
| "learning_rate": 7.685950413223142e-06, |
| "loss": 0.6348, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.08003311715192493, |
| "grad_norm": 1.6661536273928046, |
| "learning_rate": 7.961432506887054e-06, |
| "loss": 0.6263, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.08279287981233614, |
| "grad_norm": 1.2692327811914956, |
| "learning_rate": 8.236914600550965e-06, |
| "loss": 0.6294, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.08555264247274734, |
| "grad_norm": 1.3688551805953726, |
| "learning_rate": 8.512396694214877e-06, |
| "loss": 0.6268, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.08831240513315855, |
| "grad_norm": 1.3470949176246652, |
| "learning_rate": 8.787878787878788e-06, |
| "loss": 0.6247, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.09107216779356975, |
| "grad_norm": 1.136787185698281, |
| "learning_rate": 9.063360881542702e-06, |
| "loss": 0.619, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.09383193045398096, |
| "grad_norm": 1.1593643608420736, |
| "learning_rate": 9.338842975206613e-06, |
| "loss": 0.6262, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.09659169311439217, |
| "grad_norm": 1.1708021309308343, |
| "learning_rate": 9.614325068870525e-06, |
| "loss": 0.6241, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.09935145577480337, |
| "grad_norm": 1.1394208261112413, |
| "learning_rate": 9.889807162534436e-06, |
| "loss": 0.6247, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.10211121843521458, |
| "grad_norm": 1.243982861233763, |
| "learning_rate": 9.999916470583429e-06, |
| "loss": 0.6223, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.10487098109562577, |
| "grad_norm": 1.0950836157642558, |
| "learning_rate": 9.999406023144514e-06, |
| "loss": 0.6172, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.10763074375603698, |
| "grad_norm": 1.1223718246628647, |
| "learning_rate": 9.998431580815314e-06, |
| "loss": 0.6178, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.11039050641644818, |
| "grad_norm": 1.2801215739716154, |
| "learning_rate": 9.996993234033826e-06, |
| "loss": 0.6187, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.11315026907685939, |
| "grad_norm": 1.1553035959732303, |
| "learning_rate": 9.995091116293022e-06, |
| "loss": 0.6195, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.11591003173727059, |
| "grad_norm": 1.163353982074495, |
| "learning_rate": 9.992725404128452e-06, |
| "loss": 0.6162, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.1186697943976818, |
| "grad_norm": 1.0553152985762932, |
| "learning_rate": 9.989896317101873e-06, |
| "loss": 0.6075, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.12142955705809301, |
| "grad_norm": 1.061674081293769, |
| "learning_rate": 9.986604117780861e-06, |
| "loss": 0.6157, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.1241893197185042, |
| "grad_norm": 1.0865591957408132, |
| "learning_rate": 9.982849111714445e-06, |
| "loss": 0.6147, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.12694908237891542, |
| "grad_norm": 1.039091554305776, |
| "learning_rate": 9.978631647404755e-06, |
| "loss": 0.6089, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.12970884503932661, |
| "grad_norm": 1.1565672987905995, |
| "learning_rate": 9.973952116274664e-06, |
| "loss": 0.6132, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.1324686076997378, |
| "grad_norm": 1.141107303982074, |
| "learning_rate": 9.968810952631473e-06, |
| "loss": 0.6077, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.13522837036014904, |
| "grad_norm": 1.1858850784459474, |
| "learning_rate": 9.9632086336266e-06, |
| "loss": 0.6116, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.13798813302056023, |
| "grad_norm": 1.0965450940865213, |
| "learning_rate": 9.957145679211288e-06, |
| "loss": 0.6131, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.13798813302056023, |
| "eval_loss": 0.6089209318161011, |
| "eval_runtime": 49.9455, |
| "eval_samples_per_second": 58.624, |
| "eval_steps_per_second": 3.664, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.14074789568097143, |
| "grad_norm": 1.076032821296556, |
| "learning_rate": 9.95062265208836e-06, |
| "loss": 0.606, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.14350765834138265, |
| "grad_norm": 1.0371790631769926, |
| "learning_rate": 9.943640157659984e-06, |
| "loss": 0.6071, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.14626742100179385, |
| "grad_norm": 1.0166966431904405, |
| "learning_rate": 9.936198843971493e-06, |
| "loss": 0.6021, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.14902718366220505, |
| "grad_norm": 1.3219793417334544, |
| "learning_rate": 9.928299401651236e-06, |
| "loss": 0.5981, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.15178694632261625, |
| "grad_norm": 1.0389071117117548, |
| "learning_rate": 9.919942563846482e-06, |
| "loss": 0.6005, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.15454670898302747, |
| "grad_norm": 1.0246893171620508, |
| "learning_rate": 9.911129106155375e-06, |
| "loss": 0.6046, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.15730647164343867, |
| "grad_norm": 1.0445099757660354, |
| "learning_rate": 9.901859846554955e-06, |
| "loss": 0.6053, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.16006623430384986, |
| "grad_norm": 1.064106651611097, |
| "learning_rate": 9.892135645325238e-06, |
| "loss": 0.5984, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.16282599696426106, |
| "grad_norm": 1.0462870772071302, |
| "learning_rate": 9.881957404969373e-06, |
| "loss": 0.5978, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.16558575962467229, |
| "grad_norm": 1.0000683469984344, |
| "learning_rate": 9.871326070129885e-06, |
| "loss": 0.6094, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.16834552228508348, |
| "grad_norm": 0.9977951236432633, |
| "learning_rate": 9.860242627500994e-06, |
| "loss": 0.5974, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.17110528494549468, |
| "grad_norm": 4.8609383522246405, |
| "learning_rate": 9.848708105737049e-06, |
| "loss": 0.598, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.1738650476059059, |
| "grad_norm": 1.0005981539720457, |
| "learning_rate": 9.836723575357056e-06, |
| "loss": 0.5988, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.1766248102663171, |
| "grad_norm": 1.0638070414980612, |
| "learning_rate": 9.824290148645322e-06, |
| "loss": 0.5957, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.1793845729267283, |
| "grad_norm": 1.0406376026434327, |
| "learning_rate": 9.811408979548219e-06, |
| "loss": 0.591, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.1821443355871395, |
| "grad_norm": 1.07624118905585, |
| "learning_rate": 9.7980812635671e-06, |
| "loss": 0.5948, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.18490409824755072, |
| "grad_norm": 1.049133255644815, |
| "learning_rate": 9.784308237647329e-06, |
| "loss": 0.5954, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.18766386090796192, |
| "grad_norm": 1.0143310556909468, |
| "learning_rate": 9.770091180063489e-06, |
| "loss": 0.596, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.1904236235683731, |
| "grad_norm": 1.0508407539599804, |
| "learning_rate": 9.755431410300743e-06, |
| "loss": 0.5871, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.19318338622878434, |
| "grad_norm": 1.2937338319267142, |
| "learning_rate": 9.740330288932379e-06, |
| "loss": 0.5948, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.19594314888919553, |
| "grad_norm": 1.0318704519944248, |
| "learning_rate": 9.724789217493514e-06, |
| "loss": 0.5898, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.19870291154960673, |
| "grad_norm": 0.9605096481233693, |
| "learning_rate": 9.708809638351048e-06, |
| "loss": 0.5925, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.20146267421001793, |
| "grad_norm": 1.0371369229103382, |
| "learning_rate": 9.692393034569776e-06, |
| "loss": 0.5917, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.20422243687042915, |
| "grad_norm": 0.97367889824368, |
| "learning_rate": 9.675540929774751e-06, |
| "loss": 0.5911, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.20698219953084035, |
| "grad_norm": 0.9914313419740074, |
| "learning_rate": 9.658254888009877e-06, |
| "loss": 0.5862, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.20974196219125155, |
| "grad_norm": 0.9822682480321101, |
| "learning_rate": 9.64053651359275e-06, |
| "loss": 0.5871, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.21250172485166274, |
| "grad_norm": 1.001214024547401, |
| "learning_rate": 9.622387450965758e-06, |
| "loss": 0.5851, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.21526148751207397, |
| "grad_norm": 0.9608822251160972, |
| "learning_rate": 9.603809384543472e-06, |
| "loss": 0.5937, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.21802125017248516, |
| "grad_norm": 1.003270353697929, |
| "learning_rate": 9.584804038556297e-06, |
| "loss": 0.5834, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.22078101283289636, |
| "grad_norm": 1.0623920143347612, |
| "learning_rate": 9.56537317689046e-06, |
| "loss": 0.5903, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.2235407754933076, |
| "grad_norm": 0.9490080408477698, |
| "learning_rate": 9.5455186029243e-06, |
| "loss": 0.5872, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.22630053815371878, |
| "grad_norm": 0.944874262643946, |
| "learning_rate": 9.525242159360897e-06, |
| "loss": 0.5888, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.22906030081412998, |
| "grad_norm": 0.9755675580087483, |
| "learning_rate": 9.504545728057046e-06, |
| "loss": 0.5822, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.23182006347454118, |
| "grad_norm": 0.9892178439036814, |
| "learning_rate": 9.483431229848607e-06, |
| "loss": 0.5847, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.2345798261349524, |
| "grad_norm": 1.3473013721931224, |
| "learning_rate": 9.461900624372233e-06, |
| "loss": 0.5763, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.2373395887953636, |
| "grad_norm": 1.1368854612150283, |
| "learning_rate": 9.439955909883493e-06, |
| "loss": 0.5815, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.2400993514557748, |
| "grad_norm": 1.0103375787124518, |
| "learning_rate": 9.417599123071417e-06, |
| "loss": 0.5838, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.24285911411618602, |
| "grad_norm": 1.0432433126635938, |
| "learning_rate": 9.39483233886946e-06, |
| "loss": 0.5785, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.24561887677659722, |
| "grad_norm": 1.1847765522018567, |
| "learning_rate": 9.371657670262947e-06, |
| "loss": 0.5742, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.2483786394370084, |
| "grad_norm": 0.9243729444752771, |
| "learning_rate": 9.348077268092951e-06, |
| "loss": 0.5763, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.2511384020974196, |
| "grad_norm": 1.3733927460707915, |
| "learning_rate": 9.324093320856679e-06, |
| "loss": 0.5777, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.25389816475783084, |
| "grad_norm": 1.098267274159629, |
| "learning_rate": 9.299708054504355e-06, |
| "loss": 0.5781, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.256657927418242, |
| "grad_norm": 0.940463162328034, |
| "learning_rate": 9.274923732232635e-06, |
| "loss": 0.5813, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.25941769007865323, |
| "grad_norm": 0.936901862695869, |
| "learning_rate": 9.249742654274554e-06, |
| "loss": 0.576, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.26217745273906445, |
| "grad_norm": 0.9160606268594537, |
| "learning_rate": 9.224167157686044e-06, |
| "loss": 0.5756, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.2649372153994756, |
| "grad_norm": 0.902157881662283, |
| "learning_rate": 9.198199616129033e-06, |
| "loss": 0.5718, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.26769697805988685, |
| "grad_norm": 0.9251021746256725, |
| "learning_rate": 9.171842439651143e-06, |
| "loss": 0.5722, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.27045674072029807, |
| "grad_norm": 0.9049498340659763, |
| "learning_rate": 9.145098074462012e-06, |
| "loss": 0.577, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.27321650338070924, |
| "grad_norm": 1.0139196145776603, |
| "learning_rate": 9.117969002706267e-06, |
| "loss": 0.5738, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.27597626604112047, |
| "grad_norm": 0.9648188962102778, |
| "learning_rate": 9.090457742233152e-06, |
| "loss": 0.5714, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.27597626604112047, |
| "eval_loss": 0.5767696499824524, |
| "eval_runtime": 48.8421, |
| "eval_samples_per_second": 59.948, |
| "eval_steps_per_second": 3.747, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.2787360287015317, |
| "grad_norm": 0.994421928125615, |
| "learning_rate": 9.062566846362843e-06, |
| "loss": 0.5746, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.28149579136194286, |
| "grad_norm": 0.9627083740767366, |
| "learning_rate": 9.034298903649485e-06, |
| "loss": 0.5731, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.2842555540223541, |
| "grad_norm": 0.9143645503640928, |
| "learning_rate": 9.005656537640942e-06, |
| "loss": 0.574, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.2870153166827653, |
| "grad_norm": 1.0384970550188914, |
| "learning_rate": 8.976642406635295e-06, |
| "loss": 0.5725, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.2897750793431765, |
| "grad_norm": 0.9669853254759909, |
| "learning_rate": 8.947259203434147e-06, |
| "loss": 0.5737, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.2925348420035877, |
| "grad_norm": 0.965075042317439, |
| "learning_rate": 8.917509655092691e-06, |
| "loss": 0.5723, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.29529460466399887, |
| "grad_norm": 0.9720003134521186, |
| "learning_rate": 8.887396522666608e-06, |
| "loss": 0.566, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.2980543673244101, |
| "grad_norm": 1.0399992404707865, |
| "learning_rate": 8.85692260095582e-06, |
| "loss": 0.5683, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.3008141299848213, |
| "grad_norm": 1.0679186267160017, |
| "learning_rate": 8.826090718245112e-06, |
| "loss": 0.5661, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.3035738926452325, |
| "grad_norm": 0.9643258915292706, |
| "learning_rate": 8.794903736041622e-06, |
| "loss": 0.5717, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.3063336553056437, |
| "grad_norm": 0.9136180676106652, |
| "learning_rate": 8.763364548809279e-06, |
| "loss": 0.5658, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.30909341796605494, |
| "grad_norm": 0.9512590668833296, |
| "learning_rate": 8.731476083700154e-06, |
| "loss": 0.5679, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.3118531806264661, |
| "grad_norm": 0.8874320591828697, |
| "learning_rate": 8.699241300282806e-06, |
| "loss": 0.5684, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.31461294328687733, |
| "grad_norm": 0.8922502034099749, |
| "learning_rate": 8.666663190267596e-06, |
| "loss": 0.5621, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.31737270594728856, |
| "grad_norm": 1.6201987068583557, |
| "learning_rate": 8.633744777229029e-06, |
| "loss": 0.569, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.3201324686076997, |
| "grad_norm": 0.9202299543891491, |
| "learning_rate": 8.600489116325128e-06, |
| "loss": 0.5678, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.32289223126811095, |
| "grad_norm": 1.007642513849446, |
| "learning_rate": 8.566899294013901e-06, |
| "loss": 0.5696, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.3256519939285221, |
| "grad_norm": 0.9837333100849195, |
| "learning_rate": 8.53297842776687e-06, |
| "loss": 0.5655, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.32841175658893335, |
| "grad_norm": 0.9874781114751451, |
| "learning_rate": 8.498729665779751e-06, |
| "loss": 0.5653, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.33117151924934457, |
| "grad_norm": 0.9230978368811615, |
| "learning_rate": 8.464156186680262e-06, |
| "loss": 0.5665, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.33393128190975574, |
| "grad_norm": 0.8844537320977038, |
| "learning_rate": 8.429261199233114e-06, |
| "loss": 0.5633, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.33669104457016696, |
| "grad_norm": 0.8743072862956215, |
| "learning_rate": 8.394047942042215e-06, |
| "loss": 0.5648, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.3394508072305782, |
| "grad_norm": 0.9571103987457272, |
| "learning_rate": 8.358519683250087e-06, |
| "loss": 0.5628, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.34221056989098936, |
| "grad_norm": 0.9109271244972179, |
| "learning_rate": 8.322679720234553e-06, |
| "loss": 0.5665, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.3449703325514006, |
| "grad_norm": 0.9812446794213074, |
| "learning_rate": 8.286531379302703e-06, |
| "loss": 0.5615, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.3477300952118118, |
| "grad_norm": 0.9841621501852277, |
| "learning_rate": 8.25007801538218e-06, |
| "loss": 0.5626, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.350489857872223, |
| "grad_norm": 0.9203538923510197, |
| "learning_rate": 8.21332301170982e-06, |
| "loss": 0.5645, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.3532496205326342, |
| "grad_norm": 1.0479420531686792, |
| "learning_rate": 8.17626977951764e-06, |
| "loss": 0.5622, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.3560093831930454, |
| "grad_norm": 0.9787960431652523, |
| "learning_rate": 8.138921757716245e-06, |
| "loss": 0.5581, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.3587691458534566, |
| "grad_norm": 0.9530332373900086, |
| "learning_rate": 8.101282412575673e-06, |
| "loss": 0.5618, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.3615289085138678, |
| "grad_norm": 1.0048234119883286, |
| "learning_rate": 8.063355237403672e-06, |
| "loss": 0.5555, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.364288671174279, |
| "grad_norm": 0.9557255183834055, |
| "learning_rate": 8.0251437522215e-06, |
| "loss": 0.5588, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.3670484338346902, |
| "grad_norm": 0.8586667700859966, |
| "learning_rate": 7.986651503437233e-06, |
| "loss": 0.5638, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.36980819649510144, |
| "grad_norm": 0.910151087145653, |
| "learning_rate": 7.947882063516612e-06, |
| "loss": 0.561, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.3725679591555126, |
| "grad_norm": 1.055588127662, |
| "learning_rate": 7.908839030651488e-06, |
| "loss": 0.558, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.37532772181592383, |
| "grad_norm": 0.9158993525267682, |
| "learning_rate": 7.869526028425878e-06, |
| "loss": 0.5587, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.37808748447633506, |
| "grad_norm": 0.9374100988045402, |
| "learning_rate": 7.829946705479654e-06, |
| "loss": 0.5584, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.3808472471367462, |
| "grad_norm": 0.9220438433292831, |
| "learning_rate": 7.790104735169915e-06, |
| "loss": 0.5576, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.38360700979715745, |
| "grad_norm": 0.987945504497949, |
| "learning_rate": 7.750003815230062e-06, |
| "loss": 0.5558, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.3863667724575687, |
| "grad_norm": 0.9218901797189658, |
| "learning_rate": 7.70964766742662e-06, |
| "loss": 0.5611, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.38912653511797984, |
| "grad_norm": 0.964522451572826, |
| "learning_rate": 7.669040037213795e-06, |
| "loss": 0.5559, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.39188629777839107, |
| "grad_norm": 0.8962914985254243, |
| "learning_rate": 7.628184693385896e-06, |
| "loss": 0.5601, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.39464606043880224, |
| "grad_norm": 1.0900746038682583, |
| "learning_rate": 7.587085427727523e-06, |
| "loss": 0.5555, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.39740582309921346, |
| "grad_norm": 0.8919597256872009, |
| "learning_rate": 7.54574605466166e-06, |
| "loss": 0.554, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.4001655857596247, |
| "grad_norm": 0.937111912961035, |
| "learning_rate": 7.504170410895668e-06, |
| "loss": 0.5576, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.40292534842003586, |
| "grad_norm": 0.9458249661943667, |
| "learning_rate": 7.462362355065189e-06, |
| "loss": 0.5531, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.4056851110804471, |
| "grad_norm": 0.9326011700869207, |
| "learning_rate": 7.420325767376026e-06, |
| "loss": 0.5536, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.4084448737408583, |
| "grad_norm": 1.0186083692765828, |
| "learning_rate": 7.378064549244031e-06, |
| "loss": 0.5512, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.4112046364012695, |
| "grad_norm": 1.2256145317450289, |
| "learning_rate": 7.335582622933e-06, |
| "loss": 0.5601, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.4139643990616807, |
| "grad_norm": 0.8742188645170639, |
| "learning_rate": 7.292883931190667e-06, |
| "loss": 0.5524, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.4139643990616807, |
| "eval_loss": 0.5561984777450562, |
| "eval_runtime": 49.4336, |
| "eval_samples_per_second": 59.231, |
| "eval_steps_per_second": 3.702, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.4167241617220919, |
| "grad_norm": 1.2121655526375013, |
| "learning_rate": 7.249972436882756e-06, |
| "loss": 0.5542, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.4194839243825031, |
| "grad_norm": 0.9357492214275386, |
| "learning_rate": 7.206852122625203e-06, |
| "loss": 0.5524, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.4222436870429143, |
| "grad_norm": 0.9433912153799177, |
| "learning_rate": 7.163526990414522e-06, |
| "loss": 0.5542, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.4250034497033255, |
| "grad_norm": 1.1073143594605637, |
| "learning_rate": 7.120001061256387e-06, |
| "loss": 0.5525, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.4277632123637367, |
| "grad_norm": 0.9302028752309601, |
| "learning_rate": 7.076278374792429e-06, |
| "loss": 0.5541, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.43052297502414794, |
| "grad_norm": 0.9606456277899846, |
| "learning_rate": 7.032362988925332e-06, |
| "loss": 0.5553, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.4332827376845591, |
| "grad_norm": 0.9251439989697735, |
| "learning_rate": 6.9882589794422105e-06, |
| "loss": 0.5489, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.43604250034497033, |
| "grad_norm": 0.9323884014701264, |
| "learning_rate": 6.943970439636336e-06, |
| "loss": 0.5494, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.43880226300538155, |
| "grad_norm": 1.3158449326861952, |
| "learning_rate": 6.899501479927242e-06, |
| "loss": 0.5484, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.4415620256657927, |
| "grad_norm": 0.8775790961507058, |
| "learning_rate": 6.8548562274792325e-06, |
| "loss": 0.5453, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.44432178832620395, |
| "grad_norm": 0.9478549256736298, |
| "learning_rate": 6.81003882581834e-06, |
| "loss": 0.5482, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.4470815509866152, |
| "grad_norm": 0.9723342894189143, |
| "learning_rate": 6.765053434447769e-06, |
| "loss": 0.5487, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.44984131364702634, |
| "grad_norm": 0.974749959194784, |
| "learning_rate": 6.7199042284618484e-06, |
| "loss": 0.5505, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.45260107630743757, |
| "grad_norm": 0.9255671399322642, |
| "learning_rate": 6.674595398158541e-06, |
| "loss": 0.5493, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.4553608389678488, |
| "grad_norm": 0.9465763497145431, |
| "learning_rate": 6.629131148650543e-06, |
| "loss": 0.5444, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.45812060162825996, |
| "grad_norm": 0.9292686081943923, |
| "learning_rate": 6.583515699475009e-06, |
| "loss": 0.5496, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.4608803642886712, |
| "grad_norm": 1.108323416936679, |
| "learning_rate": 6.537753284201935e-06, |
| "loss": 0.5487, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.46364012694908235, |
| "grad_norm": 0.9042384031621171, |
| "learning_rate": 6.491848150041242e-06, |
| "loss": 0.5477, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.4663998896094936, |
| "grad_norm": 0.850329034972769, |
| "learning_rate": 6.4458045574485875e-06, |
| "loss": 0.5519, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.4691596522699048, |
| "grad_norm": 0.9490531745735549, |
| "learning_rate": 6.399626779729959e-06, |
| "loss": 0.5439, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.471919414930316, |
| "grad_norm": 0.9352622534171564, |
| "learning_rate": 6.353319102645069e-06, |
| "loss": 0.5465, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.4746791775907272, |
| "grad_norm": 0.9191680266246547, |
| "learning_rate": 6.306885824009585e-06, |
| "loss": 0.547, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.4774389402511384, |
| "grad_norm": 0.8597116149610817, |
| "learning_rate": 6.260331253296259e-06, |
| "loss": 0.5495, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.4801987029115496, |
| "grad_norm": 0.929502998942292, |
| "learning_rate": 6.213659711234958e-06, |
| "loss": 0.5481, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.4829584655719608, |
| "grad_norm": 0.9675046561306946, |
| "learning_rate": 6.1668755294116655e-06, |
| "loss": 0.5468, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.48571822823237204, |
| "grad_norm": 0.9073333178919157, |
| "learning_rate": 6.119983049866456e-06, |
| "loss": 0.5438, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.4884779908927832, |
| "grad_norm": 0.9109565773453232, |
| "learning_rate": 6.072986624690516e-06, |
| "loss": 0.5478, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.49123775355319443, |
| "grad_norm": 0.8955027642684514, |
| "learning_rate": 6.025890615622233e-06, |
| "loss": 0.5415, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.4939975162136056, |
| "grad_norm": 0.8648546291027864, |
| "learning_rate": 5.97869939364237e-06, |
| "loss": 0.5411, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.4967572788740168, |
| "grad_norm": 0.9523614008291538, |
| "learning_rate": 5.9314173385683986e-06, |
| "loss": 0.5453, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.49951704153442805, |
| "grad_norm": 0.9383064266388607, |
| "learning_rate": 5.884048838648017e-06, |
| "loss": 0.5384, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.5022768041948392, |
| "grad_norm": 0.9021624786717238, |
| "learning_rate": 5.836598290151866e-06, |
| "loss": 0.5456, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.5050365668552504, |
| "grad_norm": 0.9088357380698893, |
| "learning_rate": 5.789070096965514e-06, |
| "loss": 0.5426, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.5077963295156617, |
| "grad_norm": 0.922862032580903, |
| "learning_rate": 5.741468670180737e-06, |
| "loss": 0.5434, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.5105560921760729, |
| "grad_norm": 1.4926573419318607, |
| "learning_rate": 5.6937984276861195e-06, |
| "loss": 0.5441, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.513315854836484, |
| "grad_norm": 0.8580464247033338, |
| "learning_rate": 5.646063793757028e-06, |
| "loss": 0.5413, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.5160756174968952, |
| "grad_norm": 0.8836024449571962, |
| "learning_rate": 5.598269198645008e-06, |
| "loss": 0.5457, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.5188353801573065, |
| "grad_norm": 0.8951578721801148, |
| "learning_rate": 5.550419078166594e-06, |
| "loss": 0.5452, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.5215951428177177, |
| "grad_norm": 0.9955192132312739, |
| "learning_rate": 5.502517873291632e-06, |
| "loss": 0.5467, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.5243549054781289, |
| "grad_norm": 1.1988654703686727, |
| "learning_rate": 5.454570029731115e-06, |
| "loss": 0.5413, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.5271146681385401, |
| "grad_norm": 0.8839237136665734, |
| "learning_rate": 5.406579997524567e-06, |
| "loss": 0.5411, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.5298744307989512, |
| "grad_norm": 0.9047836025418462, |
| "learning_rate": 5.358552230627044e-06, |
| "loss": 0.539, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.5326341934593625, |
| "grad_norm": 0.875356281556686, |
| "learning_rate": 5.310491186495757e-06, |
| "loss": 0.5429, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.5353939561197737, |
| "grad_norm": 0.9167892739496736, |
| "learning_rate": 5.262401325676378e-06, |
| "loss": 0.5384, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.5381537187801849, |
| "grad_norm": 0.9525976300842104, |
| "learning_rate": 5.214287111389057e-06, |
| "loss": 0.5362, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.5409134814405961, |
| "grad_norm": 0.885862636010332, |
| "learning_rate": 5.166153009114188e-06, |
| "loss": 0.5378, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.5436732441010073, |
| "grad_norm": 0.9013246458040829, |
| "learning_rate": 5.1180034861779685e-06, |
| "loss": 0.5399, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.5464330067614185, |
| "grad_norm": 0.9017690594117211, |
| "learning_rate": 5.069843011337789e-06, |
| "loss": 0.5412, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.5491927694218297, |
| "grad_norm": 0.9063217665862804, |
| "learning_rate": 5.0216760543674855e-06, |
| "loss": 0.5415, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.5519525320822409, |
| "grad_norm": 0.8411472472974292, |
| "learning_rate": 4.973507085642502e-06, |
| "loss": 0.537, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.5519525320822409, |
| "eval_loss": 0.540687620639801, |
| "eval_runtime": 48.7225, |
| "eval_samples_per_second": 60.095, |
| "eval_steps_per_second": 3.756, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.5547122947426522, |
| "grad_norm": 0.8884146283634436, |
| "learning_rate": 4.92534057572499e-06, |
| "loss": 0.5406, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.5574720574030634, |
| "grad_norm": 0.8914897264034541, |
| "learning_rate": 4.8771809949489056e-06, |
| "loss": 0.5385, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.5602318200634745, |
| "grad_norm": 0.936304320970765, |
| "learning_rate": 4.829032813005103e-06, |
| "loss": 0.542, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.5629915827238857, |
| "grad_norm": 0.9310122808691009, |
| "learning_rate": 4.780900498526515e-06, |
| "loss": 0.5365, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.5657513453842969, |
| "grad_norm": 0.9518666584186597, |
| "learning_rate": 4.732788518673418e-06, |
| "loss": 0.5353, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.5685111080447082, |
| "grad_norm": 0.9510071761987918, |
| "learning_rate": 4.684701338718825e-06, |
| "loss": 0.5317, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.5712708707051194, |
| "grad_norm": 0.9432991523879628, |
| "learning_rate": 4.636643421634075e-06, |
| "loss": 0.5371, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.5740306333655306, |
| "grad_norm": 0.8853811386420297, |
| "learning_rate": 4.588619227674619e-06, |
| "loss": 0.5365, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.5767903960259417, |
| "grad_norm": 0.9290294341366813, |
| "learning_rate": 4.540633213966064e-06, |
| "loss": 0.5334, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.579550158686353, |
| "grad_norm": 0.9327668087903835, |
| "learning_rate": 4.492689834090508e-06, |
| "loss": 0.5341, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.5823099213467642, |
| "grad_norm": 1.3176571259191412, |
| "learning_rate": 4.444793537673204e-06, |
| "loss": 0.5306, |
| "step": 2110 |
| }, |
| { |
| "epoch": 0.5850696840071754, |
| "grad_norm": 0.9485574547153368, |
| "learning_rate": 4.396948769969587e-06, |
| "loss": 0.5379, |
| "step": 2120 |
| }, |
| { |
| "epoch": 0.5878294466675866, |
| "grad_norm": 0.951959101524703, |
| "learning_rate": 4.3491599714527115e-06, |
| "loss": 0.5348, |
| "step": 2130 |
| }, |
| { |
| "epoch": 0.5905892093279977, |
| "grad_norm": 0.8736078038744113, |
| "learning_rate": 4.301431577401136e-06, |
| "loss": 0.5323, |
| "step": 2140 |
| }, |
| { |
| "epoch": 0.593348971988409, |
| "grad_norm": 0.8979881907613325, |
| "learning_rate": 4.253768017487275e-06, |
| "loss": 0.5389, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.5961087346488202, |
| "grad_norm": 0.9371779488004714, |
| "learning_rate": 4.206173715366289e-06, |
| "loss": 0.5343, |
| "step": 2160 |
| }, |
| { |
| "epoch": 0.5988684973092314, |
| "grad_norm": 0.9563875289878281, |
| "learning_rate": 4.1586530882655226e-06, |
| "loss": 0.5352, |
| "step": 2170 |
| }, |
| { |
| "epoch": 0.6016282599696426, |
| "grad_norm": 0.9747450112581771, |
| "learning_rate": 4.111210546574545e-06, |
| "loss": 0.534, |
| "step": 2180 |
| }, |
| { |
| "epoch": 0.6043880226300539, |
| "grad_norm": 0.9351891869474309, |
| "learning_rate": 4.063850493435808e-06, |
| "loss": 0.5298, |
| "step": 2190 |
| }, |
| { |
| "epoch": 0.607147785290465, |
| "grad_norm": 0.8980425451702817, |
| "learning_rate": 4.0165773243360105e-06, |
| "loss": 0.5338, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.6099075479508762, |
| "grad_norm": 0.9224682367333663, |
| "learning_rate": 3.96939542669814e-06, |
| "loss": 0.5306, |
| "step": 2210 |
| }, |
| { |
| "epoch": 0.6126673106112874, |
| "grad_norm": 0.8491438680952316, |
| "learning_rate": 3.922309179474279e-06, |
| "loss": 0.5306, |
| "step": 2220 |
| }, |
| { |
| "epoch": 0.6154270732716987, |
| "grad_norm": 0.9550543136397366, |
| "learning_rate": 3.875322952739196e-06, |
| "loss": 0.5348, |
| "step": 2230 |
| }, |
| { |
| "epoch": 0.6181868359321099, |
| "grad_norm": 0.8390020003541692, |
| "learning_rate": 3.828441107284755e-06, |
| "loss": 0.5343, |
| "step": 2240 |
| }, |
| { |
| "epoch": 0.620946598592521, |
| "grad_norm": 0.892945059961568, |
| "learning_rate": 3.7816679942151945e-06, |
| "loss": 0.523, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.6237063612529322, |
| "grad_norm": 0.9210249115479134, |
| "learning_rate": 3.7350079545433014e-06, |
| "loss": 0.5279, |
| "step": 2260 |
| }, |
| { |
| "epoch": 0.6264661239133434, |
| "grad_norm": 0.9570746572961814, |
| "learning_rate": 3.6884653187875193e-06, |
| "loss": 0.5293, |
| "step": 2270 |
| }, |
| { |
| "epoch": 0.6292258865737547, |
| "grad_norm": 0.8534492122472642, |
| "learning_rate": 3.642044406570031e-06, |
| "loss": 0.5256, |
| "step": 2280 |
| }, |
| { |
| "epoch": 0.6319856492341659, |
| "grad_norm": 0.8732168431838571, |
| "learning_rate": 3.595749526215862e-06, |
| "loss": 0.5328, |
| "step": 2290 |
| }, |
| { |
| "epoch": 0.6347454118945771, |
| "grad_norm": 0.8962727112777485, |
| "learning_rate": 3.549584974353018e-06, |
| "loss": 0.5336, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.6375051745549882, |
| "grad_norm": 0.8730073595406601, |
| "learning_rate": 3.5035550355137156e-06, |
| "loss": 0.5318, |
| "step": 2310 |
| }, |
| { |
| "epoch": 0.6402649372153995, |
| "grad_norm": 0.9162256052176202, |
| "learning_rate": 3.457663981736739e-06, |
| "loss": 0.5346, |
| "step": 2320 |
| }, |
| { |
| "epoch": 0.6430246998758107, |
| "grad_norm": 0.8555722066919174, |
| "learning_rate": 3.411916072170946e-06, |
| "loss": 0.5318, |
| "step": 2330 |
| }, |
| { |
| "epoch": 0.6457844625362219, |
| "grad_norm": 0.8407652045731654, |
| "learning_rate": 3.3663155526799827e-06, |
| "loss": 0.5306, |
| "step": 2340 |
| }, |
| { |
| "epoch": 0.6485442251966331, |
| "grad_norm": 0.876217339145573, |
| "learning_rate": 3.3208666554482216e-06, |
| "loss": 0.5291, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.6513039878570442, |
| "grad_norm": 0.897083649993733, |
| "learning_rate": 3.275573598587969e-06, |
| "loss": 0.5229, |
| "step": 2360 |
| }, |
| { |
| "epoch": 0.6540637505174555, |
| "grad_norm": 0.7995897085360415, |
| "learning_rate": 3.230440585747991e-06, |
| "loss": 0.527, |
| "step": 2370 |
| }, |
| { |
| "epoch": 0.6568235131778667, |
| "grad_norm": 0.9014903537041772, |
| "learning_rate": 3.185471805723365e-06, |
| "loss": 0.5261, |
| "step": 2380 |
| }, |
| { |
| "epoch": 0.6595832758382779, |
| "grad_norm": 0.856492997294296, |
| "learning_rate": 3.140671432066719e-06, |
| "loss": 0.5308, |
| "step": 2390 |
| }, |
| { |
| "epoch": 0.6623430384986891, |
| "grad_norm": 0.888225217505466, |
| "learning_rate": 3.096043622700888e-06, |
| "loss": 0.5294, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.6651028011591004, |
| "grad_norm": 0.9531940854665482, |
| "learning_rate": 3.0515925195330148e-06, |
| "loss": 0.5283, |
| "step": 2410 |
| }, |
| { |
| "epoch": 0.6678625638195115, |
| "grad_norm": 1.2539256252009454, |
| "learning_rate": 3.0073222480701354e-06, |
| "loss": 0.5275, |
| "step": 2420 |
| }, |
| { |
| "epoch": 0.6706223264799227, |
| "grad_norm": 0.9137548818626098, |
| "learning_rate": 2.9632369170362977e-06, |
| "loss": 0.5235, |
| "step": 2430 |
| }, |
| { |
| "epoch": 0.6733820891403339, |
| "grad_norm": 1.6494495739553416, |
| "learning_rate": 2.9193406179912297e-06, |
| "loss": 0.5271, |
| "step": 2440 |
| }, |
| { |
| "epoch": 0.6761418518007452, |
| "grad_norm": 0.8590622849369576, |
| "learning_rate": 2.875637424950595e-06, |
| "loss": 0.522, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.6789016144611564, |
| "grad_norm": 0.9186749142898574, |
| "learning_rate": 2.832131394007891e-06, |
| "loss": 0.5221, |
| "step": 2460 |
| }, |
| { |
| "epoch": 0.6816613771215675, |
| "grad_norm": 0.9090102002494171, |
| "learning_rate": 2.788826562958e-06, |
| "loss": 0.5289, |
| "step": 2470 |
| }, |
| { |
| "epoch": 0.6844211397819787, |
| "grad_norm": 1.0599147672457048, |
| "learning_rate": 2.745726950922444e-06, |
| "loss": 0.5241, |
| "step": 2480 |
| }, |
| { |
| "epoch": 0.6871809024423899, |
| "grad_norm": 0.8881000835172156, |
| "learning_rate": 2.7028365579763606e-06, |
| "loss": 0.5298, |
| "step": 2490 |
| }, |
| { |
| "epoch": 0.6899406651028012, |
| "grad_norm": 0.8992940208666876, |
| "learning_rate": 2.6601593647772696e-06, |
| "loss": 0.5282, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.6899406651028012, |
| "eval_loss": 0.5283246636390686, |
| "eval_runtime": 51.449, |
| "eval_samples_per_second": 56.911, |
| "eval_steps_per_second": 3.557, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.6927004277632124, |
| "grad_norm": 0.9769386150489252, |
| "learning_rate": 2.6176993321956185e-06, |
| "loss": 0.5253, |
| "step": 2510 |
| }, |
| { |
| "epoch": 0.6954601904236236, |
| "grad_norm": 0.9191890198831723, |
| "learning_rate": 2.5754604009471786e-06, |
| "loss": 0.5229, |
| "step": 2520 |
| }, |
| { |
| "epoch": 0.6982199530840347, |
| "grad_norm": 0.8879005162140707, |
| "learning_rate": 2.533446491227305e-06, |
| "loss": 0.5247, |
| "step": 2530 |
| }, |
| { |
| "epoch": 0.700979715744446, |
| "grad_norm": 0.9109715281747162, |
| "learning_rate": 2.491661502347106e-06, |
| "loss": 0.5218, |
| "step": 2540 |
| }, |
| { |
| "epoch": 0.7037394784048572, |
| "grad_norm": 0.8969876737597425, |
| "learning_rate": 2.4501093123715395e-06, |
| "loss": 0.5238, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.7064992410652684, |
| "grad_norm": 0.9123939456680997, |
| "learning_rate": 2.408793777759504e-06, |
| "loss": 0.5234, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.7092590037256796, |
| "grad_norm": 0.8767338790553574, |
| "learning_rate": 2.3677187330059084e-06, |
| "loss": 0.5247, |
| "step": 2570 |
| }, |
| { |
| "epoch": 0.7120187663860909, |
| "grad_norm": 0.9605486388857951, |
| "learning_rate": 2.3268879902857978e-06, |
| "loss": 0.5194, |
| "step": 2580 |
| }, |
| { |
| "epoch": 0.714778529046502, |
| "grad_norm": 0.9895359728969572, |
| "learning_rate": 2.2863053391005462e-06, |
| "loss": 0.5236, |
| "step": 2590 |
| }, |
| { |
| "epoch": 0.7175382917069132, |
| "grad_norm": 0.957733912839131, |
| "learning_rate": 2.245974545926152e-06, |
| "loss": 0.5179, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.7202980543673244, |
| "grad_norm": 0.9601154451397089, |
| "learning_rate": 2.205899353863665e-06, |
| "loss": 0.5224, |
| "step": 2610 |
| }, |
| { |
| "epoch": 0.7230578170277356, |
| "grad_norm": 0.9084441231449814, |
| "learning_rate": 2.166083482291801e-06, |
| "loss": 0.5254, |
| "step": 2620 |
| }, |
| { |
| "epoch": 0.7258175796881469, |
| "grad_norm": 1.0841389588139543, |
| "learning_rate": 2.1265306265217382e-06, |
| "loss": 0.5214, |
| "step": 2630 |
| }, |
| { |
| "epoch": 0.728577342348558, |
| "grad_norm": 1.451146075094232, |
| "learning_rate": 2.0872444574541574e-06, |
| "loss": 0.5298, |
| "step": 2640 |
| }, |
| { |
| "epoch": 0.7313371050089692, |
| "grad_norm": 0.9009140743152368, |
| "learning_rate": 2.048228621238547e-06, |
| "loss": 0.5209, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.7340968676693804, |
| "grad_norm": 0.8967623610111647, |
| "learning_rate": 2.0094867389347982e-06, |
| "loss": 0.5248, |
| "step": 2660 |
| }, |
| { |
| "epoch": 0.7368566303297917, |
| "grad_norm": 0.9029822617189499, |
| "learning_rate": 1.971022406177142e-06, |
| "loss": 0.5227, |
| "step": 2670 |
| }, |
| { |
| "epoch": 0.7396163929902029, |
| "grad_norm": 0.9009675358712146, |
| "learning_rate": 1.932839192840436e-06, |
| "loss": 0.522, |
| "step": 2680 |
| }, |
| { |
| "epoch": 0.7423761556506141, |
| "grad_norm": 0.8816707412986489, |
| "learning_rate": 1.8949406427088407e-06, |
| "loss": 0.5223, |
| "step": 2690 |
| }, |
| { |
| "epoch": 0.7451359183110252, |
| "grad_norm": 0.8605202814409114, |
| "learning_rate": 1.8573302731469255e-06, |
| "loss": 0.5241, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.7478956809714364, |
| "grad_norm": 0.9612849756128029, |
| "learning_rate": 1.820011574773221e-06, |
| "loss": 0.5264, |
| "step": 2710 |
| }, |
| { |
| "epoch": 0.7506554436318477, |
| "grad_norm": 0.8793842539758803, |
| "learning_rate": 1.7829880111362486e-06, |
| "loss": 0.5181, |
| "step": 2720 |
| }, |
| { |
| "epoch": 0.7534152062922589, |
| "grad_norm": 1.2081068952668312, |
| "learning_rate": 1.746263018393079e-06, |
| "loss": 0.5292, |
| "step": 2730 |
| }, |
| { |
| "epoch": 0.7561749689526701, |
| "grad_norm": 0.8860653822887559, |
| "learning_rate": 1.7098400049904163e-06, |
| "loss": 0.5219, |
| "step": 2740 |
| }, |
| { |
| "epoch": 0.7589347316130812, |
| "grad_norm": 0.8901286343549357, |
| "learning_rate": 1.6737223513482591e-06, |
| "loss": 0.5259, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.7616944942734925, |
| "grad_norm": 0.9081512876330592, |
| "learning_rate": 1.6379134095461673e-06, |
| "loss": 0.5171, |
| "step": 2760 |
| }, |
| { |
| "epoch": 0.7644542569339037, |
| "grad_norm": 0.854771397996212, |
| "learning_rate": 1.6024165030121542e-06, |
| "loss": 0.521, |
| "step": 2770 |
| }, |
| { |
| "epoch": 0.7672140195943149, |
| "grad_norm": 0.9522518174715512, |
| "learning_rate": 1.567234926214236e-06, |
| "loss": 0.5149, |
| "step": 2780 |
| }, |
| { |
| "epoch": 0.7699737822547261, |
| "grad_norm": 0.855171965732468, |
| "learning_rate": 1.5323719443546785e-06, |
| "loss": 0.5217, |
| "step": 2790 |
| }, |
| { |
| "epoch": 0.7727335449151373, |
| "grad_norm": 0.846551026714045, |
| "learning_rate": 1.4978307930669483e-06, |
| "loss": 0.5174, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.7754933075755485, |
| "grad_norm": 0.9109328397942944, |
| "learning_rate": 1.4636146781154164e-06, |
| "loss": 0.5262, |
| "step": 2810 |
| }, |
| { |
| "epoch": 0.7782530702359597, |
| "grad_norm": 0.8916543609749644, |
| "learning_rate": 1.4297267750978277e-06, |
| "loss": 0.5258, |
| "step": 2820 |
| }, |
| { |
| "epoch": 0.7810128328963709, |
| "grad_norm": 0.839376705554234, |
| "learning_rate": 1.3961702291505791e-06, |
| "loss": 0.5233, |
| "step": 2830 |
| }, |
| { |
| "epoch": 0.7837725955567821, |
| "grad_norm": 0.9530767315659909, |
| "learning_rate": 1.3629481546568163e-06, |
| "loss": 0.5153, |
| "step": 2840 |
| }, |
| { |
| "epoch": 0.7865323582171934, |
| "grad_norm": 0.9433520691735623, |
| "learning_rate": 1.3300636349573882e-06, |
| "loss": 0.5169, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.7892921208776045, |
| "grad_norm": 0.8633774616839734, |
| "learning_rate": 1.2975197220646807e-06, |
| "loss": 0.5211, |
| "step": 2860 |
| }, |
| { |
| "epoch": 0.7920518835380157, |
| "grad_norm": 0.8830161483028809, |
| "learning_rate": 1.2653194363793642e-06, |
| "loss": 0.5169, |
| "step": 2870 |
| }, |
| { |
| "epoch": 0.7948116461984269, |
| "grad_norm": 0.905868276953366, |
| "learning_rate": 1.2334657664100614e-06, |
| "loss": 0.5167, |
| "step": 2880 |
| }, |
| { |
| "epoch": 0.7975714088588381, |
| "grad_norm": 0.8402578283356454, |
| "learning_rate": 1.2019616684959934e-06, |
| "loss": 0.5207, |
| "step": 2890 |
| }, |
| { |
| "epoch": 0.8003311715192494, |
| "grad_norm": 1.0070145612320756, |
| "learning_rate": 1.1708100665325967e-06, |
| "loss": 0.5176, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.8030909341796606, |
| "grad_norm": 0.8631246623697408, |
| "learning_rate": 1.1400138517001564e-06, |
| "loss": 0.5147, |
| "step": 2910 |
| }, |
| { |
| "epoch": 0.8058506968400717, |
| "grad_norm": 0.8746596064396687, |
| "learning_rate": 1.1095758821954788e-06, |
| "loss": 0.5212, |
| "step": 2920 |
| }, |
| { |
| "epoch": 0.8086104595004829, |
| "grad_norm": 0.8440888494194767, |
| "learning_rate": 1.0794989829666197e-06, |
| "loss": 0.5154, |
| "step": 2930 |
| }, |
| { |
| "epoch": 0.8113702221608942, |
| "grad_norm": 0.8369885923181033, |
| "learning_rate": 1.049785945450697e-06, |
| "loss": 0.516, |
| "step": 2940 |
| }, |
| { |
| "epoch": 0.8141299848213054, |
| "grad_norm": 1.5997002594596106, |
| "learning_rate": 1.0204395273148277e-06, |
| "loss": 0.5197, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.8168897474817166, |
| "grad_norm": 0.8873807689235719, |
| "learning_rate": 9.914624522001792e-07, |
| "loss": 0.5232, |
| "step": 2960 |
| }, |
| { |
| "epoch": 0.8196495101421277, |
| "grad_norm": 0.8697811520680214, |
| "learning_rate": 9.628574094691945e-07, |
| "loss": 0.5139, |
| "step": 2970 |
| }, |
| { |
| "epoch": 0.822409272802539, |
| "grad_norm": 0.8378502425029508, |
| "learning_rate": 9.346270539559882e-07, |
| "loss": 0.52, |
| "step": 2980 |
| }, |
| { |
| "epoch": 0.8251690354629502, |
| "grad_norm": 0.8849196731646373, |
| "learning_rate": 9.067740057199514e-07, |
| "loss": 0.518, |
| "step": 2990 |
| }, |
| { |
| "epoch": 0.8279287981233614, |
| "grad_norm": 0.8664792504731555, |
| "learning_rate": 8.793008498025879e-07, |
| "loss": 0.5155, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.8279287981233614, |
| "eval_loss": 0.5207065939903259, |
| "eval_runtime": 49.1882, |
| "eval_samples_per_second": 59.526, |
| "eval_steps_per_second": 3.72, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.8306885607837726, |
| "grad_norm": 0.8700644465815897, |
| "learning_rate": 8.522101359875934e-07, |
| "loss": 0.5131, |
| "step": 3010 |
| }, |
| { |
| "epoch": 0.8334483234441838, |
| "grad_norm": 0.9097989981675693, |
| "learning_rate": 8.255043785642108e-07, |
| "loss": 0.5125, |
| "step": 3020 |
| }, |
| { |
| "epoch": 0.836208086104595, |
| "grad_norm": 0.8631047164247316, |
| "learning_rate": 7.991860560938786e-07, |
| "loss": 0.5234, |
| "step": 3030 |
| }, |
| { |
| "epoch": 0.8389678487650062, |
| "grad_norm": 1.2358190890284166, |
| "learning_rate": 7.732576111801982e-07, |
| "loss": 0.5176, |
| "step": 3040 |
| }, |
| { |
| "epoch": 0.8417276114254174, |
| "grad_norm": 0.8514248482095018, |
| "learning_rate": 7.477214502422281e-07, |
| "loss": 0.5223, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.8444873740858286, |
| "grad_norm": 0.8563129413552018, |
| "learning_rate": 7.225799432911557e-07, |
| "loss": 0.5219, |
| "step": 3060 |
| }, |
| { |
| "epoch": 0.8472471367462399, |
| "grad_norm": 1.6151093899813862, |
| "learning_rate": 6.978354237103264e-07, |
| "loss": 0.516, |
| "step": 3070 |
| }, |
| { |
| "epoch": 0.850006899406651, |
| "grad_norm": 2.00845392683435, |
| "learning_rate": 6.734901880386896e-07, |
| "loss": 0.5185, |
| "step": 3080 |
| }, |
| { |
| "epoch": 0.8527666620670622, |
| "grad_norm": 1.1862873972091095, |
| "learning_rate": 6.495464957576508e-07, |
| "loss": 0.5122, |
| "step": 3090 |
| }, |
| { |
| "epoch": 0.8555264247274734, |
| "grad_norm": 0.8470639388291085, |
| "learning_rate": 6.260065690813754e-07, |
| "loss": 0.5158, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.8582861873878846, |
| "grad_norm": 0.9219107715988072, |
| "learning_rate": 6.028725927505369e-07, |
| "loss": 0.5201, |
| "step": 3110 |
| }, |
| { |
| "epoch": 0.8610459500482959, |
| "grad_norm": 0.8827729120692012, |
| "learning_rate": 5.801467138295597e-07, |
| "loss": 0.5189, |
| "step": 3120 |
| }, |
| { |
| "epoch": 0.8638057127087071, |
| "grad_norm": 0.8610445224310359, |
| "learning_rate": 5.578310415073451e-07, |
| "loss": 0.515, |
| "step": 3130 |
| }, |
| { |
| "epoch": 0.8665654753691182, |
| "grad_norm": 0.870963273635567, |
| "learning_rate": 5.359276469015179e-07, |
| "loss": 0.5151, |
| "step": 3140 |
| }, |
| { |
| "epoch": 0.8693252380295294, |
| "grad_norm": 0.8396590023282043, |
| "learning_rate": 5.14438562866208e-07, |
| "loss": 0.5223, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.8720850006899407, |
| "grad_norm": 0.8992961235238134, |
| "learning_rate": 4.933657838033795e-07, |
| "loss": 0.5144, |
| "step": 3160 |
| }, |
| { |
| "epoch": 0.8748447633503519, |
| "grad_norm": 0.9038948906592709, |
| "learning_rate": 4.7271126547772773e-07, |
| "loss": 0.5165, |
| "step": 3170 |
| }, |
| { |
| "epoch": 0.8776045260107631, |
| "grad_norm": 0.8686028788473747, |
| "learning_rate": 4.524769248351718e-07, |
| "loss": 0.5157, |
| "step": 3180 |
| }, |
| { |
| "epoch": 0.8803642886711743, |
| "grad_norm": 0.8744713706157983, |
| "learning_rate": 4.3266463982493566e-07, |
| "loss": 0.5167, |
| "step": 3190 |
| }, |
| { |
| "epoch": 0.8831240513315854, |
| "grad_norm": 0.8953682081464681, |
| "learning_rate": 4.132762492252601e-07, |
| "loss": 0.5229, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.8858838139919967, |
| "grad_norm": 1.0500886926956579, |
| "learning_rate": 3.943135524727448e-07, |
| "loss": 0.5168, |
| "step": 3210 |
| }, |
| { |
| "epoch": 0.8886435766524079, |
| "grad_norm": 0.8891911066911182, |
| "learning_rate": 3.757783094953382e-07, |
| "loss": 0.52, |
| "step": 3220 |
| }, |
| { |
| "epoch": 0.8914033393128191, |
| "grad_norm": 0.8469428073562058, |
| "learning_rate": 3.5767224054900687e-07, |
| "loss": 0.514, |
| "step": 3230 |
| }, |
| { |
| "epoch": 0.8941631019732303, |
| "grad_norm": 0.8985979256834127, |
| "learning_rate": 3.3999702605807203e-07, |
| "loss": 0.5153, |
| "step": 3240 |
| }, |
| { |
| "epoch": 0.8969228646336415, |
| "grad_norm": 0.8740663379412131, |
| "learning_rate": 3.227543064592514e-07, |
| "loss": 0.5179, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.8996826272940527, |
| "grad_norm": 0.9253549886032714, |
| "learning_rate": 3.059456820494111e-07, |
| "loss": 0.5187, |
| "step": 3260 |
| }, |
| { |
| "epoch": 0.9024423899544639, |
| "grad_norm": 0.8657401705396123, |
| "learning_rate": 2.8957271283704067e-07, |
| "loss": 0.5188, |
| "step": 3270 |
| }, |
| { |
| "epoch": 0.9052021526148751, |
| "grad_norm": 0.8317058461633765, |
| "learning_rate": 2.736369183974685e-07, |
| "loss": 0.5161, |
| "step": 3280 |
| }, |
| { |
| "epoch": 0.9079619152752864, |
| "grad_norm": 0.8738336902844653, |
| "learning_rate": 2.5813977773183175e-07, |
| "loss": 0.515, |
| "step": 3290 |
| }, |
| { |
| "epoch": 0.9107216779356976, |
| "grad_norm": 0.9137477549092036, |
| "learning_rate": 2.430827291298099e-07, |
| "loss": 0.5198, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.9134814405961087, |
| "grad_norm": 0.8521777662643962, |
| "learning_rate": 2.2846717003613462e-07, |
| "loss": 0.5221, |
| "step": 3310 |
| }, |
| { |
| "epoch": 0.9162412032565199, |
| "grad_norm": 1.12701228509693, |
| "learning_rate": 2.1429445692089712e-07, |
| "loss": 0.5154, |
| "step": 3320 |
| }, |
| { |
| "epoch": 0.9190009659169311, |
| "grad_norm": 0.9027916863842183, |
| "learning_rate": 2.0056590515365016e-07, |
| "loss": 0.5159, |
| "step": 3330 |
| }, |
| { |
| "epoch": 0.9217607285773424, |
| "grad_norm": 0.839347519350046, |
| "learning_rate": 1.8728278888132944e-07, |
| "loss": 0.518, |
| "step": 3340 |
| }, |
| { |
| "epoch": 0.9245204912377536, |
| "grad_norm": 0.8627735870099915, |
| "learning_rate": 1.744463409100039e-07, |
| "loss": 0.5189, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.9272802538981647, |
| "grad_norm": 0.8677279393067534, |
| "learning_rate": 1.620577525904532e-07, |
| "loss": 0.5099, |
| "step": 3360 |
| }, |
| { |
| "epoch": 0.9300400165585759, |
| "grad_norm": 0.8705917700670992, |
| "learning_rate": 1.501181737076035e-07, |
| "loss": 0.5154, |
| "step": 3370 |
| }, |
| { |
| "epoch": 0.9327997792189872, |
| "grad_norm": 0.9046790697526798, |
| "learning_rate": 1.3862871237381004e-07, |
| "loss": 0.5108, |
| "step": 3380 |
| }, |
| { |
| "epoch": 0.9355595418793984, |
| "grad_norm": 0.8203655507439958, |
| "learning_rate": 1.2759043492601986e-07, |
| "loss": 0.5136, |
| "step": 3390 |
| }, |
| { |
| "epoch": 0.9383193045398096, |
| "grad_norm": 0.8949093577414237, |
| "learning_rate": 1.1700436582680108e-07, |
| "loss": 0.5192, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.9410790672002208, |
| "grad_norm": 0.8425742363423367, |
| "learning_rate": 1.068714875692628e-07, |
| "loss": 0.5116, |
| "step": 3410 |
| }, |
| { |
| "epoch": 0.943838829860632, |
| "grad_norm": 0.9030687400394452, |
| "learning_rate": 9.719274058587247e-08, |
| "loss": 0.5125, |
| "step": 3420 |
| }, |
| { |
| "epoch": 0.9465985925210432, |
| "grad_norm": 0.8753228488205499, |
| "learning_rate": 8.796902316117018e-08, |
| "loss": 0.5156, |
| "step": 3430 |
| }, |
| { |
| "epoch": 0.9493583551814544, |
| "grad_norm": 0.8396184463940205, |
| "learning_rate": 7.920119134840199e-08, |
| "loss": 0.5141, |
| "step": 3440 |
| }, |
| { |
| "epoch": 0.9521181178418656, |
| "grad_norm": 0.882889766020658, |
| "learning_rate": 7.08900588900685e-08, |
| "loss": 0.5131, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.9548778805022768, |
| "grad_norm": 0.8504174898031396, |
| "learning_rate": 6.303639714240196e-08, |
| "loss": 0.5193, |
| "step": 3460 |
| }, |
| { |
| "epoch": 0.957637643162688, |
| "grad_norm": 0.8595498992279648, |
| "learning_rate": 5.564093500377732e-08, |
| "loss": 0.515, |
| "step": 3470 |
| }, |
| { |
| "epoch": 0.9603974058230992, |
| "grad_norm": 0.8244758548316357, |
| "learning_rate": 4.870435884705982e-08, |
| "loss": 0.5135, |
| "step": 3480 |
| }, |
| { |
| "epoch": 0.9631571684835104, |
| "grad_norm": 0.9192549923625648, |
| "learning_rate": 4.22273124559075e-08, |
| "loss": 0.5165, |
| "step": 3490 |
| }, |
| { |
| "epoch": 0.9659169311439216, |
| "grad_norm": 1.8467401343991516, |
| "learning_rate": 3.621039696501794e-08, |
| "loss": 0.5106, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.9659169311439216, |
| "eval_loss": 0.5180693864822388, |
| "eval_runtime": 48.9508, |
| "eval_samples_per_second": 59.815, |
| "eval_steps_per_second": 3.738, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.9686766938043329, |
| "grad_norm": 0.9145993584663505, |
| "learning_rate": 3.065417080433841e-08, |
| "loss": 0.517, |
| "step": 3510 |
| }, |
| { |
| "epoch": 0.9714364564647441, |
| "grad_norm": 0.8570121000540161, |
| "learning_rate": 2.555914964723849e-08, |
| "loss": 0.5142, |
| "step": 3520 |
| }, |
| { |
| "epoch": 0.9741962191251552, |
| "grad_norm": 0.8678615083324009, |
| "learning_rate": 2.0925806362648847e-08, |
| "loss": 0.5198, |
| "step": 3530 |
| }, |
| { |
| "epoch": 0.9769559817855664, |
| "grad_norm": 0.8381539016578547, |
| "learning_rate": 1.6754570971176944e-08, |
| "loss": 0.5125, |
| "step": 3540 |
| }, |
| { |
| "epoch": 0.9797157444459776, |
| "grad_norm": 0.9056379885360033, |
| "learning_rate": 1.3045830605192266e-08, |
| "loss": 0.5192, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.9824755071063889, |
| "grad_norm": 0.8696346972561173, |
| "learning_rate": 9.799929472902315e-09, |
| "loss": 0.5205, |
| "step": 3560 |
| }, |
| { |
| "epoch": 0.9852352697668001, |
| "grad_norm": 3.314581793197866, |
| "learning_rate": 7.017168826401466e-09, |
| "loss": 0.5169, |
| "step": 3570 |
| }, |
| { |
| "epoch": 0.9879950324272112, |
| "grad_norm": 0.8566695475770818, |
| "learning_rate": 4.697806933715021e-09, |
| "loss": 0.5095, |
| "step": 3580 |
| }, |
| { |
| "epoch": 0.9907547950876224, |
| "grad_norm": 0.8718028954806468, |
| "learning_rate": 2.8420590548294825e-09, |
| "loss": 0.5175, |
| "step": 3590 |
| }, |
| { |
| "epoch": 0.9935145577480337, |
| "grad_norm": 0.8678996243220692, |
| "learning_rate": 1.450097421710206e-09, |
| "loss": 0.519, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.9962743204084449, |
| "grad_norm": 0.8913548884246177, |
| "learning_rate": 5.220512223219621e-10, |
| "loss": 0.5135, |
| "step": 3610 |
| }, |
| { |
| "epoch": 0.9990340830688561, |
| "grad_norm": 0.8708615108897616, |
| "learning_rate": 5.8006588636305704e-11, |
| "loss": 0.5151, |
| "step": 3620 |
| }, |
| { |
| "epoch": 1.0, |
| "step": 3624, |
| "total_flos": 1884305306419200.0, |
| "train_loss": 0.5639046454929622, |
| "train_runtime": 12600.7671, |
| "train_samples_per_second": 23.004, |
| "train_steps_per_second": 0.288 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 3624, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 200, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1884305306419200.0, |
| "train_batch_size": 5, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
