{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.7794079794079796, "eval_steps": 500, "global_step": 5400, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.005148005148005148, "grad_norm": 19.51059658759793, "learning_rate": 7.718696397941682e-07, "loss": 8.2552, "step": 10 }, { "epoch": 0.010296010296010296, "grad_norm": 19.70146681963939, "learning_rate": 1.6295025728987994e-06, "loss": 8.168, "step": 20 }, { "epoch": 0.015444015444015444, "grad_norm": 9.795523599582376, "learning_rate": 2.4871355060034306e-06, "loss": 7.5953, "step": 30 }, { "epoch": 0.02059202059202059, "grad_norm": 5.9480235723668144, "learning_rate": 3.344768439108062e-06, "loss": 6.8507, "step": 40 }, { "epoch": 0.02574002574002574, "grad_norm": 3.5099463285447223, "learning_rate": 4.202401372212693e-06, "loss": 6.3398, "step": 50 }, { "epoch": 0.03088803088803089, "grad_norm": 2.90890000793416, "learning_rate": 5.060034305317325e-06, "loss": 5.9988, "step": 60 }, { "epoch": 0.036036036036036036, "grad_norm": 3.3978017812772445, "learning_rate": 5.9176672384219555e-06, "loss": 5.6503, "step": 70 }, { "epoch": 0.04118404118404118, "grad_norm": 4.103088821630264, "learning_rate": 6.775300171526587e-06, "loss": 5.1289, "step": 80 }, { "epoch": 0.04633204633204633, "grad_norm": 4.487842621805789, "learning_rate": 7.632933104631218e-06, "loss": 4.511, "step": 90 }, { "epoch": 0.05148005148005148, "grad_norm": 7.491973587682773, "learning_rate": 8.49056603773585e-06, "loss": 3.9137, "step": 100 }, { "epoch": 0.05662805662805663, "grad_norm": 7.323430680232261, "learning_rate": 9.348198970840481e-06, "loss": 3.4784, "step": 110 }, { "epoch": 0.06177606177606178, "grad_norm": 6.619474879048035, "learning_rate": 1.0205831903945113e-05, "loss": 3.2067, "step": 120 }, { "epoch": 0.06692406692406692, "grad_norm": 3.867444417417378, "learning_rate": 1.1063464837049744e-05, "loss": 3.0386, "step": 130 }, { "epoch": 0.07207207207207207, "grad_norm": 4.117034943974073, "learning_rate": 1.1921097770154374e-05, "loss": 2.9159, "step": 140 }, { "epoch": 0.07722007722007722, "grad_norm": 6.009222032591652, "learning_rate": 1.2778730703259007e-05, "loss": 2.796, "step": 150 }, { "epoch": 0.08236808236808237, "grad_norm": 11.57505334622991, "learning_rate": 1.3636363636363637e-05, "loss": 2.7077, "step": 160 }, { "epoch": 0.08751608751608751, "grad_norm": 11.428627421387615, "learning_rate": 1.4493996569468268e-05, "loss": 2.6403, "step": 170 }, { "epoch": 0.09266409266409266, "grad_norm": 10.905546843716907, "learning_rate": 1.53516295025729e-05, "loss": 2.5735, "step": 180 }, { "epoch": 0.0978120978120978, "grad_norm": 8.661551806496005, "learning_rate": 1.620926243567753e-05, "loss": 2.5078, "step": 190 }, { "epoch": 0.10296010296010295, "grad_norm": 9.940151278870877, "learning_rate": 1.7066895368782164e-05, "loss": 2.4567, "step": 200 }, { "epoch": 0.10810810810810811, "grad_norm": 13.091142904495205, "learning_rate": 1.7924528301886792e-05, "loss": 2.3953, "step": 210 }, { "epoch": 0.11325611325611326, "grad_norm": 18.06611865183751, "learning_rate": 1.8782161234991423e-05, "loss": 2.329, "step": 220 }, { "epoch": 0.11840411840411841, "grad_norm": 16.145152244084983, "learning_rate": 1.9639794168096058e-05, "loss": 2.2691, "step": 230 }, { "epoch": 0.12355212355212356, "grad_norm": 35.1534299592601, "learning_rate": 2.0497427101200686e-05, "loss": 2.2524, "step": 240 }, { "epoch": 0.1287001287001287, "grad_norm": 14.465207787142148, "learning_rate": 2.1355060034305318e-05, "loss": 2.1974, "step": 250 }, { "epoch": 0.13384813384813385, "grad_norm": 11.353218963690644, "learning_rate": 2.2212692967409952e-05, "loss": 2.0873, "step": 260 }, { "epoch": 0.138996138996139, "grad_norm": 12.277610068118626, "learning_rate": 2.307032590051458e-05, "loss": 2.0068, "step": 270 }, { "epoch": 0.14414414414414414, "grad_norm": 10.505580540439455, "learning_rate": 2.3927958833619212e-05, "loss": 1.8888, "step": 280 }, { "epoch": 0.1492921492921493, "grad_norm": 14.960676861450345, "learning_rate": 2.4785591766723843e-05, "loss": 1.7894, "step": 290 }, { "epoch": 0.15444015444015444, "grad_norm": 8.613188849395, "learning_rate": 2.564322469982847e-05, "loss": 1.6967, "step": 300 }, { "epoch": 0.15958815958815958, "grad_norm": 6.959841307105666, "learning_rate": 2.6500857632933106e-05, "loss": 1.6019, "step": 310 }, { "epoch": 0.16473616473616473, "grad_norm": 5.864669793953841, "learning_rate": 2.7358490566037738e-05, "loss": 1.5326, "step": 320 }, { "epoch": 0.16988416988416988, "grad_norm": 8.165848406130657, "learning_rate": 2.8216123499142366e-05, "loss": 1.4842, "step": 330 }, { "epoch": 0.17503217503217502, "grad_norm": 6.747629987576382, "learning_rate": 2.9073756432247e-05, "loss": 1.426, "step": 340 }, { "epoch": 0.18018018018018017, "grad_norm": 4.897733377675901, "learning_rate": 2.9931389365351632e-05, "loss": 1.3614, "step": 350 }, { "epoch": 0.18532818532818532, "grad_norm": 5.330679969567318, "learning_rate": 3.078902229845626e-05, "loss": 1.3182, "step": 360 }, { "epoch": 0.19047619047619047, "grad_norm": 4.722018507382775, "learning_rate": 3.16466552315609e-05, "loss": 1.26, "step": 370 }, { "epoch": 0.1956241956241956, "grad_norm": 6.77470473058928, "learning_rate": 3.250428816466552e-05, "loss": 1.2128, "step": 380 }, { "epoch": 0.20077220077220076, "grad_norm": 4.902136742839717, "learning_rate": 3.3361921097770154e-05, "loss": 1.1829, "step": 390 }, { "epoch": 0.2059202059202059, "grad_norm": 5.515696505859524, "learning_rate": 3.421955403087479e-05, "loss": 1.1379, "step": 400 }, { "epoch": 0.21106821106821108, "grad_norm": 3.9113735830318603, "learning_rate": 3.507718696397942e-05, "loss": 1.1002, "step": 410 }, { "epoch": 0.21621621621621623, "grad_norm": 5.8061510583054545, "learning_rate": 3.593481989708405e-05, "loss": 1.0744, "step": 420 }, { "epoch": 0.22136422136422138, "grad_norm": 5.927584495277828, "learning_rate": 3.679245283018868e-05, "loss": 1.0304, "step": 430 }, { "epoch": 0.22651222651222652, "grad_norm": 5.920319878093127, "learning_rate": 3.765008576329331e-05, "loss": 1.0023, "step": 440 }, { "epoch": 0.23166023166023167, "grad_norm": 5.302890318730487, "learning_rate": 3.850771869639794e-05, "loss": 0.9776, "step": 450 }, { "epoch": 0.23680823680823682, "grad_norm": 5.653239199646973, "learning_rate": 3.9365351629502574e-05, "loss": 0.9482, "step": 460 }, { "epoch": 0.24195624195624196, "grad_norm": 4.852345392975872, "learning_rate": 4.0222984562607206e-05, "loss": 0.9155, "step": 470 }, { "epoch": 0.2471042471042471, "grad_norm": 4.499279652551338, "learning_rate": 4.108061749571184e-05, "loss": 0.8867, "step": 480 }, { "epoch": 0.25225225225225223, "grad_norm": 5.181168954723113, "learning_rate": 4.193825042881647e-05, "loss": 0.8756, "step": 490 }, { "epoch": 0.2574002574002574, "grad_norm": 6.255177791414999, "learning_rate": 4.27958833619211e-05, "loss": 0.8506, "step": 500 }, { "epoch": 0.2625482625482625, "grad_norm": 5.396447906024421, "learning_rate": 4.365351629502573e-05, "loss": 0.8203, "step": 510 }, { "epoch": 0.2676962676962677, "grad_norm": 7.371015885385536, "learning_rate": 4.451114922813036e-05, "loss": 0.8028, "step": 520 }, { "epoch": 0.2728442728442728, "grad_norm": 6.198537777815005, "learning_rate": 4.5368782161234994e-05, "loss": 0.7808, "step": 530 }, { "epoch": 0.277992277992278, "grad_norm": 4.485699286277638, "learning_rate": 4.6226415094339625e-05, "loss": 0.7639, "step": 540 }, { "epoch": 0.28314028314028317, "grad_norm": 6.865323396700289, "learning_rate": 4.708404802744426e-05, "loss": 0.745, "step": 550 }, { "epoch": 0.2882882882882883, "grad_norm": 4.718888462403314, "learning_rate": 4.794168096054888e-05, "loss": 0.7209, "step": 560 }, { "epoch": 0.29343629343629346, "grad_norm": 4.887504164875748, "learning_rate": 4.879931389365352e-05, "loss": 0.7015, "step": 570 }, { "epoch": 0.2985842985842986, "grad_norm": 5.372842102076737, "learning_rate": 4.965694682675815e-05, "loss": 0.6767, "step": 580 }, { "epoch": 0.30373230373230375, "grad_norm": 6.766330192939623, "learning_rate": 5.0514579759862776e-05, "loss": 0.6624, "step": 590 }, { "epoch": 0.3088803088803089, "grad_norm": 4.614837742067113, "learning_rate": 5.1372212692967414e-05, "loss": 0.6405, "step": 600 }, { "epoch": 0.31402831402831405, "grad_norm": 4.201337449433587, "learning_rate": 5.222984562607204e-05, "loss": 0.6194, "step": 610 }, { "epoch": 0.31917631917631917, "grad_norm": 5.51280432030841, "learning_rate": 5.308747855917667e-05, "loss": 0.5951, "step": 620 }, { "epoch": 0.32432432432432434, "grad_norm": 5.129963269925849, "learning_rate": 5.394511149228131e-05, "loss": 0.5777, "step": 630 }, { "epoch": 0.32947232947232946, "grad_norm": 4.412209442955385, "learning_rate": 5.480274442538593e-05, "loss": 0.5642, "step": 640 }, { "epoch": 0.33462033462033464, "grad_norm": 4.336638446576286, "learning_rate": 5.5660377358490564e-05, "loss": 0.5426, "step": 650 }, { "epoch": 0.33976833976833976, "grad_norm": 3.556283909809826, "learning_rate": 5.65180102915952e-05, "loss": 0.5112, "step": 660 }, { "epoch": 0.34491634491634493, "grad_norm": 5.225584058992361, "learning_rate": 5.737564322469983e-05, "loss": 0.5044, "step": 670 }, { "epoch": 0.35006435006435005, "grad_norm": 3.1697056899798186, "learning_rate": 5.823327615780446e-05, "loss": 0.4857, "step": 680 }, { "epoch": 0.3552123552123552, "grad_norm": 3.5258824608431003, "learning_rate": 5.90909090909091e-05, "loss": 0.4747, "step": 690 }, { "epoch": 0.36036036036036034, "grad_norm": 3.5240284197467155, "learning_rate": 5.994854202401372e-05, "loss": 0.4563, "step": 700 }, { "epoch": 0.3655083655083655, "grad_norm": 3.787588207851851, "learning_rate": 6.080617495711835e-05, "loss": 0.4382, "step": 710 }, { "epoch": 0.37065637065637064, "grad_norm": 3.28493026844208, "learning_rate": 6.166380789022299e-05, "loss": 0.4216, "step": 720 }, { "epoch": 0.3758043758043758, "grad_norm": 3.144221026231138, "learning_rate": 6.252144082332761e-05, "loss": 0.417, "step": 730 }, { "epoch": 0.38095238095238093, "grad_norm": 2.780102841682132, "learning_rate": 6.337907375643225e-05, "loss": 0.3942, "step": 740 }, { "epoch": 0.3861003861003861, "grad_norm": 3.1996820882351713, "learning_rate": 6.423670668953689e-05, "loss": 0.3787, "step": 750 }, { "epoch": 0.3912483912483912, "grad_norm": 2.8374055946677923, "learning_rate": 6.50943396226415e-05, "loss": 0.3714, "step": 760 }, { "epoch": 0.3963963963963964, "grad_norm": 3.0094036046611454, "learning_rate": 6.595197255574615e-05, "loss": 0.3606, "step": 770 }, { "epoch": 0.4015444015444015, "grad_norm": 2.6679117500690204, "learning_rate": 6.680960548885078e-05, "loss": 0.3379, "step": 780 }, { "epoch": 0.4066924066924067, "grad_norm": 3.101181469988925, "learning_rate": 6.76672384219554e-05, "loss": 0.3271, "step": 790 }, { "epoch": 0.4118404118404118, "grad_norm": 2.421653916410696, "learning_rate": 6.852487135506004e-05, "loss": 0.3132, "step": 800 }, { "epoch": 0.416988416988417, "grad_norm": 2.1110958003741174, "learning_rate": 6.938250428816467e-05, "loss": 0.3013, "step": 810 }, { "epoch": 0.42213642213642216, "grad_norm": 1.9804288224535966, "learning_rate": 7.024013722126929e-05, "loss": 0.2892, "step": 820 }, { "epoch": 0.4272844272844273, "grad_norm": 2.0965122713743725, "learning_rate": 7.109777015437394e-05, "loss": 0.2802, "step": 830 }, { "epoch": 0.43243243243243246, "grad_norm": 2.0662383619585323, "learning_rate": 7.195540308747857e-05, "loss": 0.2666, "step": 840 }, { "epoch": 0.4375804375804376, "grad_norm": 1.9080845115386118, "learning_rate": 7.281303602058319e-05, "loss": 0.2557, "step": 850 }, { "epoch": 0.44272844272844275, "grad_norm": 2.435492621224063, "learning_rate": 7.367066895368783e-05, "loss": 0.2441, "step": 860 }, { "epoch": 0.44787644787644787, "grad_norm": 2.3478908862806707, "learning_rate": 7.452830188679245e-05, "loss": 0.2302, "step": 870 }, { "epoch": 0.45302445302445304, "grad_norm": 2.033330789035785, "learning_rate": 7.538593481989708e-05, "loss": 0.222, "step": 880 }, { "epoch": 0.45817245817245816, "grad_norm": 1.8373992771294045, "learning_rate": 7.624356775300173e-05, "loss": 0.2079, "step": 890 }, { "epoch": 0.46332046332046334, "grad_norm": 1.9518420548134319, "learning_rate": 7.710120068610634e-05, "loss": 0.1979, "step": 900 }, { "epoch": 0.46846846846846846, "grad_norm": 1.6103496909038504, "learning_rate": 7.795883361921097e-05, "loss": 0.1936, "step": 910 }, { "epoch": 0.47361647361647363, "grad_norm": 1.3694934489721085, "learning_rate": 7.881646655231562e-05, "loss": 0.1841, "step": 920 }, { "epoch": 0.47876447876447875, "grad_norm": 1.2736699951465489, "learning_rate": 7.967409948542024e-05, "loss": 0.1766, "step": 930 }, { "epoch": 0.4839124839124839, "grad_norm": 2.1829922501226444, "learning_rate": 8.053173241852487e-05, "loss": 0.1675, "step": 940 }, { "epoch": 0.48906048906048905, "grad_norm": 1.659964942426995, "learning_rate": 8.138936535162951e-05, "loss": 0.161, "step": 950 }, { "epoch": 0.4942084942084942, "grad_norm": 1.5767630911351294, "learning_rate": 8.224699828473413e-05, "loss": 0.1494, "step": 960 }, { "epoch": 0.49935649935649934, "grad_norm": 1.537782758985047, "learning_rate": 8.310463121783876e-05, "loss": 0.1445, "step": 970 }, { "epoch": 0.5045045045045045, "grad_norm": 1.604101659458472, "learning_rate": 8.396226415094341e-05, "loss": 0.1402, "step": 980 }, { "epoch": 0.5096525096525096, "grad_norm": 2.256805925155236, "learning_rate": 8.481989708404803e-05, "loss": 0.1405, "step": 990 }, { "epoch": 0.5148005148005148, "grad_norm": 1.4857578289963005, "learning_rate": 8.567753001715266e-05, "loss": 0.1321, "step": 1000 }, { "epoch": 0.51994851994852, "grad_norm": 1.0326983623178083, "learning_rate": 8.65351629502573e-05, "loss": 0.1224, "step": 1010 }, { "epoch": 0.525096525096525, "grad_norm": 1.1180976317039228, "learning_rate": 8.739279588336192e-05, "loss": 0.1189, "step": 1020 }, { "epoch": 0.5302445302445302, "grad_norm": 1.1721351234336, "learning_rate": 8.825042881646655e-05, "loss": 0.1092, "step": 1030 }, { "epoch": 0.5353925353925354, "grad_norm": 0.9640093738745075, "learning_rate": 8.91080617495712e-05, "loss": 0.1021, "step": 1040 }, { "epoch": 0.5405405405405406, "grad_norm": 1.0306043643976568, "learning_rate": 8.996569468267581e-05, "loss": 0.0982, "step": 1050 }, { "epoch": 0.5456885456885456, "grad_norm": 1.4999347000799717, "learning_rate": 9.082332761578045e-05, "loss": 0.0936, "step": 1060 }, { "epoch": 0.5508365508365508, "grad_norm": 1.0452281614309868, "learning_rate": 9.168096054888509e-05, "loss": 0.0878, "step": 1070 }, { "epoch": 0.555984555984556, "grad_norm": 1.2088207649167284, "learning_rate": 9.253859348198971e-05, "loss": 0.0824, "step": 1080 }, { "epoch": 0.5611325611325612, "grad_norm": 0.6373898488496081, "learning_rate": 9.339622641509434e-05, "loss": 0.0809, "step": 1090 }, { "epoch": 0.5662805662805663, "grad_norm": 0.5917955695640877, "learning_rate": 9.425385934819899e-05, "loss": 0.0742, "step": 1100 }, { "epoch": 0.5714285714285714, "grad_norm": 0.6146276278340692, "learning_rate": 9.51114922813036e-05, "loss": 0.0706, "step": 1110 }, { "epoch": 0.5765765765765766, "grad_norm": 0.8766352471472525, "learning_rate": 9.596912521440823e-05, "loss": 0.0672, "step": 1120 }, { "epoch": 0.5817245817245817, "grad_norm": 0.6678260178109867, "learning_rate": 9.682675814751287e-05, "loss": 0.0655, "step": 1130 }, { "epoch": 0.5868725868725869, "grad_norm": 0.49817357185278055, "learning_rate": 9.76843910806175e-05, "loss": 0.0611, "step": 1140 }, { "epoch": 0.592020592020592, "grad_norm": 1.0087468424307855, "learning_rate": 9.854202401372213e-05, "loss": 0.0611, "step": 1150 }, { "epoch": 0.5971685971685972, "grad_norm": 0.5362523083451194, "learning_rate": 9.939965694682676e-05, "loss": 0.0579, "step": 1160 }, { "epoch": 0.6023166023166023, "grad_norm": 0.4534299709647709, "learning_rate": 9.99999798272238e-05, "loss": 0.0549, "step": 1170 }, { "epoch": 0.6074646074646075, "grad_norm": 0.6406776231632558, "learning_rate": 9.999962120054416e-05, "loss": 0.0534, "step": 1180 }, { "epoch": 0.6126126126126126, "grad_norm": 0.5792838241902996, "learning_rate": 9.999881429364988e-05, "loss": 0.0512, "step": 1190 }, { "epoch": 0.6177606177606177, "grad_norm": 0.4275795011357426, "learning_rate": 9.999755911377544e-05, "loss": 0.0488, "step": 1200 }, { "epoch": 0.6229086229086229, "grad_norm": 0.33615545845119615, "learning_rate": 9.999585567217439e-05, "loss": 0.0476, "step": 1210 }, { "epoch": 0.6280566280566281, "grad_norm": 0.33655008372027123, "learning_rate": 9.99937039841192e-05, "loss": 0.0444, "step": 1220 }, { "epoch": 0.6332046332046332, "grad_norm": 0.43244236090431704, "learning_rate": 9.999110406890121e-05, "loss": 0.0438, "step": 1230 }, { "epoch": 0.6383526383526383, "grad_norm": 0.41139634543075815, "learning_rate": 9.998805594983043e-05, "loss": 0.042, "step": 1240 }, { "epoch": 0.6435006435006435, "grad_norm": 0.5206972896361842, "learning_rate": 9.998455965423525e-05, "loss": 0.0431, "step": 1250 }, { "epoch": 0.6486486486486487, "grad_norm": 0.3937459395109014, "learning_rate": 9.998061521346236e-05, "loss": 0.041, "step": 1260 }, { "epoch": 0.6537966537966537, "grad_norm": 0.2963678010736684, "learning_rate": 9.997622266287629e-05, "loss": 0.0393, "step": 1270 }, { "epoch": 0.6589446589446589, "grad_norm": 0.26754815540836463, "learning_rate": 9.997138204185923e-05, "loss": 0.0376, "step": 1280 }, { "epoch": 0.6640926640926641, "grad_norm": 0.29045330821553156, "learning_rate": 9.996609339381057e-05, "loss": 0.036, "step": 1290 }, { "epoch": 0.6692406692406693, "grad_norm": 0.3089797983346258, "learning_rate": 9.996035676614663e-05, "loss": 0.0356, "step": 1300 }, { "epoch": 0.6743886743886743, "grad_norm": 0.26639822964562176, "learning_rate": 9.99541722103001e-05, "loss": 0.0349, "step": 1310 }, { "epoch": 0.6795366795366795, "grad_norm": 0.45121223083554485, "learning_rate": 9.994753978171971e-05, "loss": 0.0351, "step": 1320 }, { "epoch": 0.6846846846846847, "grad_norm": 0.3667468018122777, "learning_rate": 9.994045953986964e-05, "loss": 0.0338, "step": 1330 }, { "epoch": 0.6898326898326899, "grad_norm": 0.24841154017565364, "learning_rate": 9.993293154822901e-05, "loss": 0.0328, "step": 1340 }, { "epoch": 0.694980694980695, "grad_norm": 0.283360706082952, "learning_rate": 9.992495587429129e-05, "loss": 0.0315, "step": 1350 }, { "epoch": 0.7001287001287001, "grad_norm": 0.24413900296621255, "learning_rate": 9.991653258956381e-05, "loss": 0.0311, "step": 1360 }, { "epoch": 0.7052767052767053, "grad_norm": 0.28975076041601167, "learning_rate": 9.990766176956697e-05, "loss": 0.0303, "step": 1370 }, { "epoch": 0.7104247104247104, "grad_norm": 0.26351763932473593, "learning_rate": 9.989834349383361e-05, "loss": 0.03, "step": 1380 }, { "epoch": 0.7155727155727156, "grad_norm": 0.5331124737621773, "learning_rate": 9.988857784590836e-05, "loss": 0.0291, "step": 1390 }, { "epoch": 0.7207207207207207, "grad_norm": 0.24186391898474244, "learning_rate": 9.98783649133468e-05, "loss": 0.0284, "step": 1400 }, { "epoch": 0.7258687258687259, "grad_norm": 0.21615992134341075, "learning_rate": 9.986770478771475e-05, "loss": 0.0273, "step": 1410 }, { "epoch": 0.731016731016731, "grad_norm": 0.21264572442174073, "learning_rate": 9.985659756458741e-05, "loss": 0.0267, "step": 1420 }, { "epoch": 0.7361647361647362, "grad_norm": 0.2320920564652625, "learning_rate": 9.984504334354847e-05, "loss": 0.0262, "step": 1430 }, { "epoch": 0.7413127413127413, "grad_norm": 0.2342099433653173, "learning_rate": 9.983304222818929e-05, "loss": 0.0252, "step": 1440 }, { "epoch": 0.7464607464607464, "grad_norm": 0.240834162772985, "learning_rate": 9.9820594326108e-05, "loss": 0.0245, "step": 1450 }, { "epoch": 0.7516087516087516, "grad_norm": 0.2161277200650416, "learning_rate": 9.980769974890833e-05, "loss": 0.0238, "step": 1460 }, { "epoch": 0.7567567567567568, "grad_norm": 0.24019983928841215, "learning_rate": 9.979435861219888e-05, "loss": 0.0233, "step": 1470 }, { "epoch": 0.7619047619047619, "grad_norm": 0.2970865898808262, "learning_rate": 9.978057103559188e-05, "loss": 0.0226, "step": 1480 }, { "epoch": 0.767052767052767, "grad_norm": 0.24780480301331292, "learning_rate": 9.976633714270226e-05, "loss": 0.0221, "step": 1490 }, { "epoch": 0.7722007722007722, "grad_norm": 0.25727613523882426, "learning_rate": 9.975165706114639e-05, "loss": 0.0214, "step": 1500 }, { "epoch": 0.7773487773487774, "grad_norm": 0.24359971967254024, "learning_rate": 9.973653092254113e-05, "loss": 0.0209, "step": 1510 }, { "epoch": 0.7824967824967825, "grad_norm": 0.2067746721639629, "learning_rate": 9.97209588625024e-05, "loss": 0.0202, "step": 1520 }, { "epoch": 0.7876447876447876, "grad_norm": 0.19304855106218685, "learning_rate": 9.970494102064427e-05, "loss": 0.0195, "step": 1530 }, { "epoch": 0.7927927927927928, "grad_norm": 0.22693160915077307, "learning_rate": 9.968847754057742e-05, "loss": 0.019, "step": 1540 }, { "epoch": 0.797940797940798, "grad_norm": 0.23060204821027103, "learning_rate": 9.967156856990801e-05, "loss": 0.0186, "step": 1550 }, { "epoch": 0.803088803088803, "grad_norm": 2.572239383211685, "learning_rate": 9.965421426023631e-05, "loss": 0.1139, "step": 1560 }, { "epoch": 0.8082368082368082, "grad_norm": 0.5437887934638203, "learning_rate": 9.963641476715541e-05, "loss": 0.0343, "step": 1570 }, { "epoch": 0.8133848133848134, "grad_norm": 0.23796250360876128, "learning_rate": 9.961817025024974e-05, "loss": 0.0235, "step": 1580 }, { "epoch": 0.8185328185328186, "grad_norm": 0.1764844020309923, "learning_rate": 9.959948087309362e-05, "loss": 0.0195, "step": 1590 }, { "epoch": 0.8236808236808236, "grad_norm": 0.17189189871306304, "learning_rate": 9.95803468032499e-05, "loss": 0.0177, "step": 1600 }, { "epoch": 0.8288288288288288, "grad_norm": 0.17751312710898323, "learning_rate": 9.956076821226836e-05, "loss": 0.0171, "step": 1610 }, { "epoch": 0.833976833976834, "grad_norm": 0.16437143945406296, "learning_rate": 9.954074527568426e-05, "loss": 0.0161, "step": 1620 }, { "epoch": 0.8391248391248392, "grad_norm": 0.15832400304822497, "learning_rate": 9.952027817301667e-05, "loss": 0.0157, "step": 1630 }, { "epoch": 0.8442728442728443, "grad_norm": 0.16117748572605023, "learning_rate": 9.949936708776691e-05, "loss": 0.0148, "step": 1640 }, { "epoch": 0.8494208494208494, "grad_norm": 0.16984391534714557, "learning_rate": 9.947801220741695e-05, "loss": 0.0143, "step": 1650 }, { "epoch": 0.8545688545688546, "grad_norm": 0.1696223065684197, "learning_rate": 9.945621372342762e-05, "loss": 0.0137, "step": 1660 }, { "epoch": 0.8597168597168597, "grad_norm": 0.1800035242900872, "learning_rate": 9.943397183123697e-05, "loss": 0.0133, "step": 1670 }, { "epoch": 0.8648648648648649, "grad_norm": 0.15249239519006993, "learning_rate": 9.941128673025854e-05, "loss": 0.0133, "step": 1680 }, { "epoch": 0.87001287001287, "grad_norm": 0.1551659031233157, "learning_rate": 9.938815862387953e-05, "loss": 0.0132, "step": 1690 }, { "epoch": 0.8751608751608752, "grad_norm": 0.14348975402940561, "learning_rate": 9.936458771945891e-05, "loss": 0.0131, "step": 1700 }, { "epoch": 0.8803088803088803, "grad_norm": 0.140601933336413, "learning_rate": 9.934057422832572e-05, "loss": 0.0123, "step": 1710 }, { "epoch": 0.8854568854568855, "grad_norm": 0.1275666846983677, "learning_rate": 9.931611836577703e-05, "loss": 0.012, "step": 1720 }, { "epoch": 0.8906048906048906, "grad_norm": 0.13327069275308084, "learning_rate": 9.92912203510761e-05, "loss": 0.012, "step": 1730 }, { "epoch": 0.8957528957528957, "grad_norm": 0.1658754837807317, "learning_rate": 9.926588040745035e-05, "loss": 0.0117, "step": 1740 }, { "epoch": 0.9009009009009009, "grad_norm": 0.14050126640493185, "learning_rate": 9.924009876208944e-05, "loss": 0.0116, "step": 1750 }, { "epoch": 0.9060489060489061, "grad_norm": 0.1413904681709387, "learning_rate": 9.921387564614313e-05, "loss": 0.0111, "step": 1760 }, { "epoch": 0.9111969111969112, "grad_norm": 0.13671787855200412, "learning_rate": 9.918721129471927e-05, "loss": 0.0113, "step": 1770 }, { "epoch": 0.9163449163449163, "grad_norm": 0.1300287060008007, "learning_rate": 9.916010594688173e-05, "loss": 0.011, "step": 1780 }, { "epoch": 0.9214929214929215, "grad_norm": 0.12272142768731135, "learning_rate": 9.913255984564814e-05, "loss": 0.0104, "step": 1790 }, { "epoch": 0.9266409266409267, "grad_norm": 0.13585333911080333, "learning_rate": 9.910457323798783e-05, "loss": 0.0101, "step": 1800 }, { "epoch": 0.9317889317889317, "grad_norm": 0.1365832152868149, "learning_rate": 9.907614637481957e-05, "loss": 0.0104, "step": 1810 }, { "epoch": 0.9369369369369369, "grad_norm": 0.13886380017866182, "learning_rate": 9.904727951100926e-05, "loss": 0.0099, "step": 1820 }, { "epoch": 0.9420849420849421, "grad_norm": 0.13559410476056236, "learning_rate": 9.901797290536775e-05, "loss": 0.0098, "step": 1830 }, { "epoch": 0.9472329472329473, "grad_norm": 0.1380086457171486, "learning_rate": 9.898822682064848e-05, "loss": 0.0098, "step": 1840 }, { "epoch": 0.9523809523809523, "grad_norm": 0.6177427987927422, "learning_rate": 9.895804152354507e-05, "loss": 0.0098, "step": 1850 }, { "epoch": 0.9575289575289575, "grad_norm": 0.508965354576064, "learning_rate": 9.892741728468902e-05, "loss": 0.0154, "step": 1860 }, { "epoch": 0.9626769626769627, "grad_norm": 0.1734010623999643, "learning_rate": 9.889635437864719e-05, "loss": 0.0125, "step": 1870 }, { "epoch": 0.9678249678249679, "grad_norm": 0.15703615270146173, "learning_rate": 9.886485308391944e-05, "loss": 0.0104, "step": 1880 }, { "epoch": 0.972972972972973, "grad_norm": 0.14238560159782349, "learning_rate": 9.883291368293605e-05, "loss": 0.0093, "step": 1890 }, { "epoch": 0.9781209781209781, "grad_norm": 0.11593480562868673, "learning_rate": 9.880053646205523e-05, "loss": 0.0088, "step": 1900 }, { "epoch": 0.9832689832689833, "grad_norm": 0.1250311748465019, "learning_rate": 9.87677217115605e-05, "loss": 0.0087, "step": 1910 }, { "epoch": 0.9884169884169884, "grad_norm": 0.10915721832834366, "learning_rate": 9.873446972565816e-05, "loss": 0.0084, "step": 1920 }, { "epoch": 0.9935649935649936, "grad_norm": 0.11826713704360893, "learning_rate": 9.870078080247464e-05, "loss": 0.0084, "step": 1930 }, { "epoch": 0.9987129987129987, "grad_norm": 0.13132056186197924, "learning_rate": 9.866665524405377e-05, "loss": 0.0082, "step": 1940 }, { "epoch": 1.0036036036036036, "grad_norm": 0.11858168124629603, "learning_rate": 9.86320933563541e-05, "loss": 0.0077, "step": 1950 }, { "epoch": 1.0087516087516089, "grad_norm": 0.1237462850171375, "learning_rate": 9.859709544924624e-05, "loss": 0.0076, "step": 1960 }, { "epoch": 1.013899613899614, "grad_norm": 0.1224220645410722, "learning_rate": 9.85616618365099e-05, "loss": 0.0075, "step": 1970 }, { "epoch": 1.019047619047619, "grad_norm": 0.13079765571571764, "learning_rate": 9.852579283583126e-05, "loss": 0.0075, "step": 1980 }, { "epoch": 1.0241956241956243, "grad_norm": 0.11011593961405876, "learning_rate": 9.848948876880002e-05, "loss": 0.0073, "step": 1990 }, { "epoch": 1.0293436293436293, "grad_norm": 0.10429883248159245, "learning_rate": 9.845274996090655e-05, "loss": 0.007, "step": 2000 }, { "epoch": 1.0344916344916344, "grad_norm": 0.10883414844159683, "learning_rate": 9.841557674153897e-05, "loss": 0.0069, "step": 2010 }, { "epoch": 1.0396396396396397, "grad_norm": 0.10985446797058561, "learning_rate": 9.837796944398014e-05, "loss": 0.0068, "step": 2020 }, { "epoch": 1.0447876447876447, "grad_norm": 0.10929404720847335, "learning_rate": 9.833992840540481e-05, "loss": 0.0069, "step": 2030 }, { "epoch": 1.04993564993565, "grad_norm": 0.10611541757073792, "learning_rate": 9.830145396687641e-05, "loss": 0.0068, "step": 2040 }, { "epoch": 1.055083655083655, "grad_norm": 0.09681663896670493, "learning_rate": 9.826254647334419e-05, "loss": 0.0066, "step": 2050 }, { "epoch": 1.0602316602316602, "grad_norm": 0.10666808939263758, "learning_rate": 9.822320627363997e-05, "loss": 0.0064, "step": 2060 }, { "epoch": 1.0653796653796654, "grad_norm": 0.11874311376321617, "learning_rate": 9.81834337204751e-05, "loss": 0.0065, "step": 2070 }, { "epoch": 1.0705276705276705, "grad_norm": 0.10330219941649453, "learning_rate": 9.814322917043721e-05, "loss": 0.0065, "step": 2080 }, { "epoch": 1.0756756756756758, "grad_norm": 0.10518179079688333, "learning_rate": 9.810259298398717e-05, "loss": 0.0063, "step": 2090 }, { "epoch": 1.0808236808236809, "grad_norm": 0.11196650081811249, "learning_rate": 9.806152552545567e-05, "loss": 0.0061, "step": 2100 }, { "epoch": 1.085971685971686, "grad_norm": 0.10391144612302242, "learning_rate": 9.80200271630401e-05, "loss": 0.0063, "step": 2110 }, { "epoch": 1.0911196911196912, "grad_norm": 0.10437009383950677, "learning_rate": 9.797809826880118e-05, "loss": 0.006, "step": 2120 }, { "epoch": 1.0962676962676963, "grad_norm": 0.15742816962279707, "learning_rate": 9.793573921865961e-05, "loss": 0.0061, "step": 2130 }, { "epoch": 1.1014157014157013, "grad_norm": 1.2391996261097045, "learning_rate": 9.789295039239277e-05, "loss": 0.009, "step": 2140 }, { "epoch": 1.1065637065637066, "grad_norm": 0.4258283630396214, "learning_rate": 9.784973217363123e-05, "loss": 0.0167, "step": 2150 }, { "epoch": 1.1117117117117117, "grad_norm": 0.22704282205376575, "learning_rate": 9.78060849498554e-05, "loss": 0.011, "step": 2160 }, { "epoch": 1.1168597168597167, "grad_norm": 0.2903222759151862, "learning_rate": 9.776200911239196e-05, "loss": 0.0146, "step": 2170 }, { "epoch": 1.122007722007722, "grad_norm": 0.13920400223191687, "learning_rate": 9.771750505641044e-05, "loss": 0.0091, "step": 2180 }, { "epoch": 1.127155727155727, "grad_norm": 0.11920796398066835, "learning_rate": 9.767257318091961e-05, "loss": 0.0074, "step": 2190 }, { "epoch": 1.1323037323037324, "grad_norm": 0.10527324354349751, "learning_rate": 9.76272138887639e-05, "loss": 0.0068, "step": 2200 }, { "epoch": 1.1374517374517374, "grad_norm": 0.09156854644361, "learning_rate": 9.758142758661992e-05, "loss": 0.0062, "step": 2210 }, { "epoch": 1.1425997425997425, "grad_norm": 0.10256933439724165, "learning_rate": 9.753521468499261e-05, "loss": 0.0058, "step": 2220 }, { "epoch": 1.1477477477477478, "grad_norm": 0.09728084601303465, "learning_rate": 9.748857559821171e-05, "loss": 0.0059, "step": 2230 }, { "epoch": 1.1528957528957529, "grad_norm": 0.09070888908222362, "learning_rate": 9.744151074442798e-05, "loss": 0.0056, "step": 2240 }, { "epoch": 1.1580437580437581, "grad_norm": 0.08815351818450588, "learning_rate": 9.739402054560948e-05, "loss": 0.0054, "step": 2250 }, { "epoch": 1.1631917631917632, "grad_norm": 0.08362202696869882, "learning_rate": 9.734610542753775e-05, "loss": 0.0052, "step": 2260 }, { "epoch": 1.1683397683397683, "grad_norm": 0.08985409051259748, "learning_rate": 9.729776581980403e-05, "loss": 0.0054, "step": 2270 }, { "epoch": 1.1734877734877736, "grad_norm": 0.09519354418158256, "learning_rate": 9.724900215580542e-05, "loss": 0.0052, "step": 2280 }, { "epoch": 1.1786357786357786, "grad_norm": 0.09918504442751083, "learning_rate": 9.719981487274094e-05, "loss": 0.0054, "step": 2290 }, { "epoch": 1.1837837837837837, "grad_norm": 0.08289788258009483, "learning_rate": 9.715020441160764e-05, "loss": 0.0053, "step": 2300 }, { "epoch": 1.188931788931789, "grad_norm": 0.09232435159305605, "learning_rate": 9.710017121719666e-05, "loss": 0.0049, "step": 2310 }, { "epoch": 1.194079794079794, "grad_norm": 0.08339859786132367, "learning_rate": 9.704971573808921e-05, "loss": 0.0049, "step": 2320 }, { "epoch": 1.1992277992277993, "grad_norm": 0.08650033957989789, "learning_rate": 9.699883842665258e-05, "loss": 0.0048, "step": 2330 }, { "epoch": 1.2043758043758044, "grad_norm": 0.0794400511956092, "learning_rate": 9.694753973903608e-05, "loss": 0.0049, "step": 2340 }, { "epoch": 1.2095238095238094, "grad_norm": 0.0902066930501087, "learning_rate": 9.689582013516692e-05, "loss": 0.0048, "step": 2350 }, { "epoch": 1.2146718146718147, "grad_norm": 0.09349490891133622, "learning_rate": 9.68436800787461e-05, "loss": 0.0048, "step": 2360 }, { "epoch": 1.2198198198198198, "grad_norm": 0.08645959165196554, "learning_rate": 9.679112003724433e-05, "loss": 0.0047, "step": 2370 }, { "epoch": 1.224967824967825, "grad_norm": 0.10081362401668935, "learning_rate": 9.673814048189765e-05, "loss": 0.0049, "step": 2380 }, { "epoch": 1.2301158301158301, "grad_norm": 0.08712722386999017, "learning_rate": 9.668474188770342e-05, "loss": 0.0046, "step": 2390 }, { "epoch": 1.2352638352638352, "grad_norm": 0.07793748236883287, "learning_rate": 9.663092473341591e-05, "loss": 0.0045, "step": 2400 }, { "epoch": 1.2404118404118405, "grad_norm": 0.08419470824276815, "learning_rate": 9.657668950154212e-05, "loss": 0.0045, "step": 2410 }, { "epoch": 1.2455598455598456, "grad_norm": 0.08910987303942121, "learning_rate": 9.652203667833732e-05, "loss": 0.0046, "step": 2420 }, { "epoch": 1.2507078507078506, "grad_norm": 0.09000059511882977, "learning_rate": 9.646696675380084e-05, "loss": 0.0046, "step": 2430 }, { "epoch": 1.255855855855856, "grad_norm": 0.08973547907365796, "learning_rate": 9.641148022167157e-05, "loss": 0.0045, "step": 2440 }, { "epoch": 1.261003861003861, "grad_norm": 0.08752137232187301, "learning_rate": 9.635557757942356e-05, "loss": 0.0045, "step": 2450 }, { "epoch": 1.266151866151866, "grad_norm": 0.08625521978145931, "learning_rate": 9.629925932826158e-05, "loss": 0.0044, "step": 2460 }, { "epoch": 1.2712998712998713, "grad_norm": 0.09652129154021626, "learning_rate": 9.62425259731166e-05, "loss": 0.0044, "step": 2470 }, { "epoch": 1.2764478764478764, "grad_norm": 0.07738037320281026, "learning_rate": 9.61853780226413e-05, "loss": 0.0043, "step": 2480 }, { "epoch": 1.2815958815958817, "grad_norm": 0.08005866211189636, "learning_rate": 9.612781598920545e-05, "loss": 0.0042, "step": 2490 }, { "epoch": 1.2867438867438867, "grad_norm": 0.07390360062985053, "learning_rate": 9.60698403888914e-05, "loss": 0.0043, "step": 2500 }, { "epoch": 1.291891891891892, "grad_norm": 0.07290137906930576, "learning_rate": 9.601145174148938e-05, "loss": 0.0041, "step": 2510 }, { "epoch": 1.297039897039897, "grad_norm": 0.08077140524538812, "learning_rate": 9.595265057049284e-05, "loss": 0.0043, "step": 2520 }, { "epoch": 1.3021879021879021, "grad_norm": 0.071170385653506, "learning_rate": 9.589343740309382e-05, "loss": 0.0042, "step": 2530 }, { "epoch": 1.3073359073359074, "grad_norm": 0.08200339249188611, "learning_rate": 9.583381277017816e-05, "loss": 0.004, "step": 2540 }, { "epoch": 1.3124839124839125, "grad_norm": 0.07521720836461765, "learning_rate": 9.577377720632079e-05, "loss": 0.0039, "step": 2550 }, { "epoch": 1.3176319176319176, "grad_norm": 0.0901292008046331, "learning_rate": 9.57133312497809e-05, "loss": 0.0041, "step": 2560 }, { "epoch": 1.3227799227799228, "grad_norm": 0.08285967408155427, "learning_rate": 9.56524754424971e-05, "loss": 0.004, "step": 2570 }, { "epoch": 1.327927927927928, "grad_norm": 0.09219709170644368, "learning_rate": 9.55912103300826e-05, "loss": 0.0041, "step": 2580 }, { "epoch": 1.333075933075933, "grad_norm": 0.06662885941521572, "learning_rate": 9.552953646182037e-05, "loss": 0.004, "step": 2590 }, { "epoch": 1.3382239382239383, "grad_norm": 0.08022484610511026, "learning_rate": 9.546745439065805e-05, "loss": 0.004, "step": 2600 }, { "epoch": 1.3433719433719433, "grad_norm": 0.08810609300651148, "learning_rate": 9.540496467320317e-05, "loss": 0.004, "step": 2610 }, { "epoch": 1.3485199485199484, "grad_norm": 0.08370070467094351, "learning_rate": 9.534206786971802e-05, "loss": 0.004, "step": 2620 }, { "epoch": 1.3536679536679537, "grad_norm": 0.08571191252220195, "learning_rate": 9.527876454411476e-05, "loss": 0.0039, "step": 2630 }, { "epoch": 1.3588159588159587, "grad_norm": 0.08575267541892966, "learning_rate": 9.521505526395025e-05, "loss": 0.0039, "step": 2640 }, { "epoch": 1.363963963963964, "grad_norm": 0.0772459071410357, "learning_rate": 9.515094060042098e-05, "loss": 0.0039, "step": 2650 }, { "epoch": 1.369111969111969, "grad_norm": 0.09154097043649315, "learning_rate": 9.508642112835803e-05, "loss": 0.0036, "step": 2660 }, { "epoch": 1.3742599742599744, "grad_norm": 0.06979887713205823, "learning_rate": 9.502149742622186e-05, "loss": 0.0036, "step": 2670 }, { "epoch": 1.3794079794079794, "grad_norm": 0.07524766626872072, "learning_rate": 9.495617007609708e-05, "loss": 0.0037, "step": 2680 }, { "epoch": 1.3845559845559845, "grad_norm": 0.07741077734917863, "learning_rate": 9.489043966368729e-05, "loss": 0.0036, "step": 2690 }, { "epoch": 1.3897039897039898, "grad_norm": 0.07777205972495332, "learning_rate": 9.482430677830984e-05, "loss": 0.0036, "step": 2700 }, { "epoch": 1.3948519948519948, "grad_norm": 0.08027699605432952, "learning_rate": 9.475777201289049e-05, "loss": 0.0037, "step": 2710 }, { "epoch": 1.4, "grad_norm": 0.08878147480054278, "learning_rate": 9.469083596395813e-05, "loss": 0.0037, "step": 2720 }, { "epoch": 1.4051480051480052, "grad_norm": 0.0824428011807005, "learning_rate": 9.462349923163947e-05, "loss": 0.0037, "step": 2730 }, { "epoch": 1.4102960102960103, "grad_norm": 0.10261444743241076, "learning_rate": 9.455576241965357e-05, "loss": 0.0035, "step": 2740 }, { "epoch": 1.4154440154440153, "grad_norm": 0.08949325884555213, "learning_rate": 9.448762613530644e-05, "loss": 0.0036, "step": 2750 }, { "epoch": 1.4205920205920206, "grad_norm": 0.0723289853593256, "learning_rate": 9.44190909894857e-05, "loss": 0.0036, "step": 2760 }, { "epoch": 1.4257400257400257, "grad_norm": 0.08857410702799448, "learning_rate": 9.435015759665502e-05, "loss": 0.0034, "step": 2770 }, { "epoch": 1.430888030888031, "grad_norm": 0.08063193397260371, "learning_rate": 9.428082657484858e-05, "loss": 0.0035, "step": 2780 }, { "epoch": 1.436036036036036, "grad_norm": 0.07967833602585864, "learning_rate": 9.421109854566562e-05, "loss": 0.0036, "step": 2790 }, { "epoch": 1.4411840411840413, "grad_norm": 0.08170135657562853, "learning_rate": 9.414097413426477e-05, "loss": 0.0036, "step": 2800 }, { "epoch": 1.4463320463320464, "grad_norm": 0.06868763716374629, "learning_rate": 9.407045396935852e-05, "loss": 0.0033, "step": 2810 }, { "epoch": 1.4514800514800514, "grad_norm": 0.07632059161778292, "learning_rate": 9.399953868320757e-05, "loss": 0.0033, "step": 2820 }, { "epoch": 1.4566280566280567, "grad_norm": 0.07781026078770122, "learning_rate": 9.392822891161513e-05, "loss": 0.0032, "step": 2830 }, { "epoch": 1.4617760617760618, "grad_norm": 0.07814855533275045, "learning_rate": 9.38565252939212e-05, "loss": 0.0033, "step": 2840 }, { "epoch": 1.4669240669240668, "grad_norm": 0.07181980467452527, "learning_rate": 9.378442847299696e-05, "loss": 0.0034, "step": 2850 }, { "epoch": 1.4720720720720721, "grad_norm": 0.07540071489782683, "learning_rate": 9.371193909523885e-05, "loss": 0.0034, "step": 2860 }, { "epoch": 1.4772200772200772, "grad_norm": 0.07474893132191902, "learning_rate": 9.363905781056288e-05, "loss": 0.0032, "step": 2870 }, { "epoch": 1.4823680823680823, "grad_norm": 0.06752409841169732, "learning_rate": 9.356578527239876e-05, "loss": 0.0033, "step": 2880 }, { "epoch": 1.4875160875160875, "grad_norm": 0.07227853197497423, "learning_rate": 9.349212213768403e-05, "loss": 0.0033, "step": 2890 }, { "epoch": 1.4926640926640926, "grad_norm": 0.08116684705266587, "learning_rate": 9.341806906685824e-05, "loss": 0.0033, "step": 2900 }, { "epoch": 1.4978120978120977, "grad_norm": 0.07203072835662648, "learning_rate": 9.334362672385692e-05, "loss": 0.0032, "step": 2910 }, { "epoch": 1.502960102960103, "grad_norm": 0.06997940354981029, "learning_rate": 9.326879577610574e-05, "loss": 0.003, "step": 2920 }, { "epoch": 1.5081081081081082, "grad_norm": 0.07405993830075755, "learning_rate": 9.319357689451443e-05, "loss": 0.0031, "step": 2930 }, { "epoch": 1.5132561132561133, "grad_norm": 0.06306898034549013, "learning_rate": 9.311797075347084e-05, "loss": 0.0032, "step": 2940 }, { "epoch": 1.5184041184041184, "grad_norm": 0.0673006276719181, "learning_rate": 9.304197803083486e-05, "loss": 0.003, "step": 2950 }, { "epoch": 1.5235521235521237, "grad_norm": 0.06910575947165132, "learning_rate": 9.296559940793231e-05, "loss": 0.003, "step": 2960 }, { "epoch": 1.5287001287001287, "grad_norm": 0.07701591471399939, "learning_rate": 9.288883556954888e-05, "loss": 0.003, "step": 2970 }, { "epoch": 1.5338481338481338, "grad_norm": 0.07543347242871117, "learning_rate": 9.2811687203924e-05, "loss": 0.003, "step": 2980 }, { "epoch": 1.538996138996139, "grad_norm": 0.06367656554256501, "learning_rate": 9.273415500274463e-05, "loss": 0.0031, "step": 2990 }, { "epoch": 1.5441441441441441, "grad_norm": 0.07035242882025344, "learning_rate": 9.265623966113908e-05, "loss": 0.003, "step": 3000 }, { "epoch": 1.5492921492921492, "grad_norm": 0.07832135035294531, "learning_rate": 9.257794187767072e-05, "loss": 0.0031, "step": 3010 }, { "epoch": 1.5544401544401545, "grad_norm": 0.06599561720521474, "learning_rate": 9.249926235433187e-05, "loss": 0.003, "step": 3020 }, { "epoch": 1.5595881595881596, "grad_norm": 0.05267911081489989, "learning_rate": 9.242020179653732e-05, "loss": 0.0029, "step": 3030 }, { "epoch": 1.5647361647361646, "grad_norm": 0.06704594471912081, "learning_rate": 9.234076091311809e-05, "loss": 0.003, "step": 3040 }, { "epoch": 1.56988416988417, "grad_norm": 0.06460368185603607, "learning_rate": 9.226094041631515e-05, "loss": 0.0029, "step": 3050 }, { "epoch": 1.5750321750321752, "grad_norm": 0.05979445278270312, "learning_rate": 9.218074102177288e-05, "loss": 0.003, "step": 3060 }, { "epoch": 1.58018018018018, "grad_norm": 0.08016520116607714, "learning_rate": 9.210016344853276e-05, "loss": 0.0029, "step": 3070 }, { "epoch": 1.5853281853281853, "grad_norm": 0.06802817286962771, "learning_rate": 9.201920841902692e-05, "loss": 0.003, "step": 3080 }, { "epoch": 1.5904761904761906, "grad_norm": 0.06797365347482685, "learning_rate": 9.193787665907161e-05, "loss": 0.0029, "step": 3090 }, { "epoch": 1.5956241956241957, "grad_norm": 0.07102105091489752, "learning_rate": 9.185616889786073e-05, "loss": 0.0029, "step": 3100 }, { "epoch": 1.6007722007722007, "grad_norm": 0.07171846203450846, "learning_rate": 9.177408586795927e-05, "loss": 0.0029, "step": 3110 }, { "epoch": 1.605920205920206, "grad_norm": 0.06055340864499565, "learning_rate": 9.169162830529682e-05, "loss": 0.0027, "step": 3120 }, { "epoch": 1.611068211068211, "grad_norm": 0.07533819698090054, "learning_rate": 9.16087969491608e-05, "loss": 0.0028, "step": 3130 }, { "epoch": 1.6162162162162161, "grad_norm": 0.06743711306850025, "learning_rate": 9.152559254219003e-05, "loss": 0.0027, "step": 3140 }, { "epoch": 1.6213642213642214, "grad_norm": 0.06056537426381112, "learning_rate": 9.144201583036791e-05, "loss": 0.0027, "step": 3150 }, { "epoch": 1.6265122265122265, "grad_norm": 0.06508273546989307, "learning_rate": 9.135806756301591e-05, "loss": 0.0027, "step": 3160 }, { "epoch": 1.6316602316602316, "grad_norm": 0.06830147531666561, "learning_rate": 9.127374849278662e-05, "loss": 0.0028, "step": 3170 }, { "epoch": 1.6368082368082368, "grad_norm": 0.06217191434234582, "learning_rate": 9.118905937565722e-05, "loss": 0.0028, "step": 3180 }, { "epoch": 1.641956241956242, "grad_norm": 0.05780489801778332, "learning_rate": 9.110400097092256e-05, "loss": 0.0027, "step": 3190 }, { "epoch": 1.647104247104247, "grad_norm": 0.06254550151471086, "learning_rate": 9.101857404118843e-05, "loss": 0.0026, "step": 3200 }, { "epoch": 1.6522522522522523, "grad_norm": 0.05814178748214669, "learning_rate": 9.093277935236466e-05, "loss": 0.0026, "step": 3210 }, { "epoch": 1.6574002574002575, "grad_norm": 0.16718357391133418, "learning_rate": 9.08466176736583e-05, "loss": 0.0028, "step": 3220 }, { "epoch": 1.6625482625482626, "grad_norm": 2.396831293896128, "learning_rate": 9.076008977756676e-05, "loss": 0.0466, "step": 3230 }, { "epoch": 1.6676962676962677, "grad_norm": 2.7472945663163135, "learning_rate": 9.067319643987076e-05, "loss": 0.0557, "step": 3240 }, { "epoch": 1.672844272844273, "grad_norm": 0.26885148545801296, "learning_rate": 9.05859384396275e-05, "loss": 0.0304, "step": 3250 }, { "epoch": 1.677992277992278, "grad_norm": 0.12409899016396063, "learning_rate": 9.04983165591636e-05, "loss": 0.0095, "step": 3260 }, { "epoch": 1.683140283140283, "grad_norm": 0.09294953853472421, "learning_rate": 9.041033158406812e-05, "loss": 0.0055, "step": 3270 }, { "epoch": 1.6882882882882884, "grad_norm": 0.09689553530835725, "learning_rate": 9.032198430318549e-05, "loss": 0.0042, "step": 3280 }, { "epoch": 1.6934362934362934, "grad_norm": 0.09140013573445376, "learning_rate": 9.023327550860851e-05, "loss": 0.0038, "step": 3290 }, { "epoch": 1.6985842985842985, "grad_norm": 0.08116155041824391, "learning_rate": 9.014420599567113e-05, "loss": 0.0034, "step": 3300 }, { "epoch": 1.7037323037323038, "grad_norm": 0.0679507754540547, "learning_rate": 9.00547765629414e-05, "loss": 0.0032, "step": 3310 }, { "epoch": 1.7088803088803088, "grad_norm": 0.058242905790430746, "learning_rate": 8.996498801221431e-05, "loss": 0.003, "step": 3320 }, { "epoch": 1.714028314028314, "grad_norm": 0.07351637623119718, "learning_rate": 8.98748411485046e-05, "loss": 0.0029, "step": 3330 }, { "epoch": 1.7191763191763192, "grad_norm": 0.06328328422520454, "learning_rate": 8.978433678003946e-05, "loss": 0.0029, "step": 3340 }, { "epoch": 1.7243243243243245, "grad_norm": 0.07268810964129091, "learning_rate": 8.969347571825142e-05, "loss": 0.0029, "step": 3350 }, { "epoch": 1.7294723294723293, "grad_norm": 0.05741572542909016, "learning_rate": 8.960225877777095e-05, "loss": 0.0029, "step": 3360 }, { "epoch": 1.7346203346203346, "grad_norm": 0.06233872470439857, "learning_rate": 8.951068677641928e-05, "loss": 0.0027, "step": 3370 }, { "epoch": 1.7397683397683399, "grad_norm": 0.06306044970907165, "learning_rate": 8.94187605352009e-05, "loss": 0.0027, "step": 3380 }, { "epoch": 1.744916344916345, "grad_norm": 0.05866957972844069, "learning_rate": 8.932648087829643e-05, "loss": 0.0026, "step": 3390 }, { "epoch": 1.75006435006435, "grad_norm": 0.06351564249061811, "learning_rate": 8.923384863305495e-05, "loss": 0.0027, "step": 3400 }, { "epoch": 1.7552123552123553, "grad_norm": 0.06311005104924097, "learning_rate": 8.914086462998686e-05, "loss": 0.0026, "step": 3410 }, { "epoch": 1.7603603603603604, "grad_norm": 0.07175111106462256, "learning_rate": 8.904752970275624e-05, "loss": 0.0026, "step": 3420 }, { "epoch": 1.7655083655083654, "grad_norm": 0.06069156181724329, "learning_rate": 8.895384468817345e-05, "loss": 0.0025, "step": 3430 }, { "epoch": 1.7706563706563707, "grad_norm": 0.06445989759572321, "learning_rate": 8.885981042618762e-05, "loss": 0.0025, "step": 3440 }, { "epoch": 1.7758043758043758, "grad_norm": 0.06092468872567865, "learning_rate": 8.876542775987912e-05, "loss": 0.0024, "step": 3450 }, { "epoch": 1.7809523809523808, "grad_norm": 0.0610543681951457, "learning_rate": 8.867069753545202e-05, "loss": 0.0024, "step": 3460 }, { "epoch": 1.7861003861003861, "grad_norm": 0.057721016476337414, "learning_rate": 8.857562060222644e-05, "loss": 0.0023, "step": 3470 }, { "epoch": 1.7912483912483914, "grad_norm": 0.06437257969879298, "learning_rate": 8.848019781263099e-05, "loss": 0.0025, "step": 3480 }, { "epoch": 1.7963963963963963, "grad_norm": 0.06569126814994662, "learning_rate": 8.838443002219514e-05, "loss": 0.0023, "step": 3490 }, { "epoch": 1.8015444015444015, "grad_norm": 0.06150895775850208, "learning_rate": 8.82883180895415e-05, "loss": 0.0023, "step": 3500 }, { "epoch": 1.8066924066924068, "grad_norm": 0.0616857154906518, "learning_rate": 8.819186287637817e-05, "loss": 0.0023, "step": 3510 }, { "epoch": 1.811840411840412, "grad_norm": 0.09278079283374936, "learning_rate": 8.809506524749097e-05, "loss": 0.0025, "step": 3520 }, { "epoch": 1.816988416988417, "grad_norm": 0.05632880872010278, "learning_rate": 8.799792607073567e-05, "loss": 0.0024, "step": 3530 }, { "epoch": 1.8221364221364222, "grad_norm": 0.06087327975528164, "learning_rate": 8.790044621703033e-05, "loss": 0.0024, "step": 3540 }, { "epoch": 1.8272844272844273, "grad_norm": 0.05140811610791699, "learning_rate": 8.780262656034739e-05, "loss": 0.0023, "step": 3550 }, { "epoch": 1.8324324324324324, "grad_norm": 0.06674966884600453, "learning_rate": 8.770446797770577e-05, "loss": 0.0023, "step": 3560 }, { "epoch": 1.8375804375804377, "grad_norm": 0.05618543547388763, "learning_rate": 8.760597134916318e-05, "loss": 0.0022, "step": 3570 }, { "epoch": 1.8427284427284427, "grad_norm": 0.05237285038394349, "learning_rate": 8.750713755780811e-05, "loss": 0.0022, "step": 3580 }, { "epoch": 1.8478764478764478, "grad_norm": 0.058744035782622876, "learning_rate": 8.740796748975191e-05, "loss": 0.0021, "step": 3590 }, { "epoch": 1.853024453024453, "grad_norm": 0.06251312062428276, "learning_rate": 8.730846203412094e-05, "loss": 0.0021, "step": 3600 }, { "epoch": 1.8581724581724581, "grad_norm": 0.05660174173246188, "learning_rate": 8.720862208304847e-05, "loss": 0.0022, "step": 3610 }, { "epoch": 1.8633204633204632, "grad_norm": 0.05460329788545314, "learning_rate": 8.710844853166677e-05, "loss": 0.0021, "step": 3620 }, { "epoch": 1.8684684684684685, "grad_norm": 0.05920176351989063, "learning_rate": 8.70079422780991e-05, "loss": 0.0021, "step": 3630 }, { "epoch": 1.8736164736164738, "grad_norm": 0.05975993421974977, "learning_rate": 8.690710422345156e-05, "loss": 0.0021, "step": 3640 }, { "epoch": 1.8787644787644786, "grad_norm": 0.07261290977712913, "learning_rate": 8.680593527180512e-05, "loss": 0.0022, "step": 3650 }, { "epoch": 1.883912483912484, "grad_norm": 0.057238529150264346, "learning_rate": 8.67044363302074e-05, "loss": 0.0021, "step": 3660 }, { "epoch": 1.8890604890604892, "grad_norm": 0.05467721608387215, "learning_rate": 8.660260830866467e-05, "loss": 0.0021, "step": 3670 }, { "epoch": 1.8942084942084942, "grad_norm": 0.05845129114063532, "learning_rate": 8.650045212013358e-05, "loss": 0.0023, "step": 3680 }, { "epoch": 1.8993564993564993, "grad_norm": 0.07268624727169087, "learning_rate": 8.639796868051305e-05, "loss": 0.0021, "step": 3690 }, { "epoch": 1.9045045045045046, "grad_norm": 0.05876373181946239, "learning_rate": 8.629515890863596e-05, "loss": 0.0022, "step": 3700 }, { "epoch": 1.9096525096525097, "grad_norm": 0.048951820788902485, "learning_rate": 8.619202372626109e-05, "loss": 0.0021, "step": 3710 }, { "epoch": 1.9148005148005147, "grad_norm": 0.05561005341623571, "learning_rate": 8.608856405806464e-05, "loss": 0.0019, "step": 3720 }, { "epoch": 1.91994851994852, "grad_norm": 0.057100787317847096, "learning_rate": 8.598478083163208e-05, "loss": 0.002, "step": 3730 }, { "epoch": 1.925096525096525, "grad_norm": 0.059660059973863006, "learning_rate": 8.588067497744983e-05, "loss": 0.0021, "step": 3740 }, { "epoch": 1.9302445302445301, "grad_norm": 0.047168011131905614, "learning_rate": 8.57762474288968e-05, "loss": 0.0019, "step": 3750 }, { "epoch": 1.9353925353925354, "grad_norm": 0.0594860132000797, "learning_rate": 8.567149912223624e-05, "loss": 0.002, "step": 3760 }, { "epoch": 1.9405405405405407, "grad_norm": 0.060916594149851046, "learning_rate": 8.556643099660706e-05, "loss": 0.0019, "step": 3770 }, { "epoch": 1.9456885456885455, "grad_norm": 0.06191170974205215, "learning_rate": 8.546104399401567e-05, "loss": 0.0019, "step": 3780 }, { "epoch": 1.9508365508365508, "grad_norm": 0.059050634437477476, "learning_rate": 8.535533905932738e-05, "loss": 0.002, "step": 3790 }, { "epoch": 1.9559845559845561, "grad_norm": 0.05874211825594844, "learning_rate": 8.5249317140258e-05, "loss": 0.002, "step": 3800 }, { "epoch": 1.9611325611325612, "grad_norm": 0.051714056463086085, "learning_rate": 8.514297918736528e-05, "loss": 0.002, "step": 3810 }, { "epoch": 1.9662805662805662, "grad_norm": 0.05669285044549558, "learning_rate": 8.503632615404047e-05, "loss": 0.0018, "step": 3820 }, { "epoch": 1.9714285714285715, "grad_norm": 0.0527546216463499, "learning_rate": 8.492935899649971e-05, "loss": 0.0019, "step": 3830 }, { "epoch": 1.9765765765765766, "grad_norm": 0.0517620866256957, "learning_rate": 8.482207867377546e-05, "loss": 0.0019, "step": 3840 }, { "epoch": 1.9817245817245817, "grad_norm": 0.05504170017730614, "learning_rate": 8.471448614770792e-05, "loss": 0.0019, "step": 3850 }, { "epoch": 1.986872586872587, "grad_norm": 0.0531051637907689, "learning_rate": 8.460658238293644e-05, "loss": 0.0019, "step": 3860 }, { "epoch": 1.992020592020592, "grad_norm": 0.06878662684799065, "learning_rate": 8.449836834689078e-05, "loss": 0.0019, "step": 3870 }, { "epoch": 1.997168597168597, "grad_norm": 0.07389086836198688, "learning_rate": 8.438984500978253e-05, "loss": 0.002, "step": 3880 }, { "epoch": 2.002059202059202, "grad_norm": 0.04129522947698218, "learning_rate": 8.428101334459634e-05, "loss": 0.0018, "step": 3890 }, { "epoch": 2.007207207207207, "grad_norm": 0.07916532613092706, "learning_rate": 8.41718743270813e-05, "loss": 0.0015, "step": 3900 }, { "epoch": 2.0123552123552124, "grad_norm": 0.04733101682802457, "learning_rate": 8.406242893574198e-05, "loss": 0.0015, "step": 3910 }, { "epoch": 2.0175032175032177, "grad_norm": 0.0417837325535251, "learning_rate": 8.395267815182996e-05, "loss": 0.0014, "step": 3920 }, { "epoch": 2.0226512226512225, "grad_norm": 0.052954653259521055, "learning_rate": 8.384262295933477e-05, "loss": 0.0013, "step": 3930 }, { "epoch": 2.027799227799228, "grad_norm": 0.04596870277989711, "learning_rate": 8.373226434497518e-05, "loss": 0.0014, "step": 3940 }, { "epoch": 2.032947232947233, "grad_norm": 0.04826449367607423, "learning_rate": 8.362160329819038e-05, "loss": 0.0012, "step": 3950 }, { "epoch": 2.038095238095238, "grad_norm": 0.059447596021510715, "learning_rate": 8.3510640811131e-05, "loss": 0.0013, "step": 3960 }, { "epoch": 2.0432432432432432, "grad_norm": 0.04596769846728861, "learning_rate": 8.339937787865036e-05, "loss": 0.0013, "step": 3970 }, { "epoch": 2.0483912483912485, "grad_norm": 0.044589049577602416, "learning_rate": 8.328781549829543e-05, "loss": 0.0014, "step": 3980 }, { "epoch": 2.0535392535392534, "grad_norm": 0.04119681252531216, "learning_rate": 8.317595467029795e-05, "loss": 0.0014, "step": 3990 }, { "epoch": 2.0586872586872587, "grad_norm": 0.05580855727608084, "learning_rate": 8.306379639756544e-05, "loss": 0.0012, "step": 4000 }, { "epoch": 2.063835263835264, "grad_norm": 0.04940964639449694, "learning_rate": 8.29513416856722e-05, "loss": 0.0013, "step": 4010 }, { "epoch": 2.068983268983269, "grad_norm": 0.04780393223229874, "learning_rate": 8.283859154285032e-05, "loss": 0.0013, "step": 4020 }, { "epoch": 2.074131274131274, "grad_norm": 0.0581497942061673, "learning_rate": 8.272554697998063e-05, "loss": 0.0013, "step": 4030 }, { "epoch": 2.0792792792792794, "grad_norm": 0.04439803542359854, "learning_rate": 8.261220901058362e-05, "loss": 0.0013, "step": 4040 }, { "epoch": 2.084427284427284, "grad_norm": 0.03893925557971465, "learning_rate": 8.249857865081037e-05, "loss": 0.0013, "step": 4050 }, { "epoch": 2.0895752895752895, "grad_norm": 0.05448252248762289, "learning_rate": 8.238465691943344e-05, "loss": 0.0014, "step": 4060 }, { "epoch": 2.0947232947232948, "grad_norm": 0.04777291394045677, "learning_rate": 8.227044483783772e-05, "loss": 0.0013, "step": 4070 }, { "epoch": 2.0998712998713, "grad_norm": 0.04531698982390845, "learning_rate": 8.215594343001133e-05, "loss": 0.0014, "step": 4080 }, { "epoch": 2.105019305019305, "grad_norm": 0.05289715107833794, "learning_rate": 8.204115372253635e-05, "loss": 0.0014, "step": 4090 }, { "epoch": 2.11016731016731, "grad_norm": 0.05000518964787425, "learning_rate": 8.192607674457966e-05, "loss": 0.0014, "step": 4100 }, { "epoch": 2.1153153153153155, "grad_norm": 0.07423565873105893, "learning_rate": 8.181071352788371e-05, "loss": 0.0013, "step": 4110 }, { "epoch": 2.1204633204633203, "grad_norm": 0.04387669724085158, "learning_rate": 8.169506510675733e-05, "loss": 0.0013, "step": 4120 }, { "epoch": 2.1256113256113256, "grad_norm": 0.054902691601927234, "learning_rate": 8.15791325180663e-05, "loss": 0.0014, "step": 4130 }, { "epoch": 2.130759330759331, "grad_norm": 0.048346239723777266, "learning_rate": 8.146291680122427e-05, "loss": 0.0014, "step": 4140 }, { "epoch": 2.1359073359073357, "grad_norm": 0.05113058555459646, "learning_rate": 8.13464189981832e-05, "loss": 0.0012, "step": 4150 }, { "epoch": 2.141055341055341, "grad_norm": 0.04692623057151712, "learning_rate": 8.122964015342427e-05, "loss": 0.0013, "step": 4160 }, { "epoch": 2.1462033462033463, "grad_norm": 0.049807948136136, "learning_rate": 8.111258131394828e-05, "loss": 0.0013, "step": 4170 }, { "epoch": 2.1513513513513516, "grad_norm": 0.04810393599315422, "learning_rate": 8.099524352926643e-05, "loss": 0.0013, "step": 4180 }, { "epoch": 2.1564993564993564, "grad_norm": 0.04804529263899986, "learning_rate": 8.087762785139084e-05, "loss": 0.0014, "step": 4190 }, { "epoch": 2.1616473616473617, "grad_norm": 0.04394140785220738, "learning_rate": 8.07597353348251e-05, "loss": 0.0012, "step": 4200 }, { "epoch": 2.166795366795367, "grad_norm": 0.04386355702357804, "learning_rate": 8.064156703655489e-05, "loss": 0.0011, "step": 4210 }, { "epoch": 2.171943371943372, "grad_norm": 0.04119947072133049, "learning_rate": 8.052312401603847e-05, "loss": 0.0012, "step": 4220 }, { "epoch": 2.177091377091377, "grad_norm": 0.04030666528465274, "learning_rate": 8.040440733519709e-05, "loss": 0.0012, "step": 4230 }, { "epoch": 2.1822393822393824, "grad_norm": 0.057812767601739, "learning_rate": 8.028541805840561e-05, "loss": 0.0012, "step": 4240 }, { "epoch": 2.1873873873873872, "grad_norm": 0.047874075229963925, "learning_rate": 8.016615725248289e-05, "loss": 0.0015, "step": 4250 }, { "epoch": 2.1925353925353925, "grad_norm": 0.059625625401084195, "learning_rate": 8.00466259866822e-05, "loss": 0.0014, "step": 4260 }, { "epoch": 2.197683397683398, "grad_norm": 0.050376692026641155, "learning_rate": 7.992682533268175e-05, "loss": 0.0013, "step": 4270 }, { "epoch": 2.2028314028314027, "grad_norm": 0.048462338863960994, "learning_rate": 7.980675636457487e-05, "loss": 0.0014, "step": 4280 }, { "epoch": 2.207979407979408, "grad_norm": 0.04835528405206845, "learning_rate": 7.968642015886058e-05, "loss": 0.0013, "step": 4290 }, { "epoch": 2.2131274131274132, "grad_norm": 0.039780282771479415, "learning_rate": 7.95658177944339e-05, "loss": 0.0011, "step": 4300 }, { "epoch": 2.218275418275418, "grad_norm": 0.04452063196388458, "learning_rate": 7.944495035257605e-05, "loss": 0.0012, "step": 4310 }, { "epoch": 2.2234234234234234, "grad_norm": 0.0456971239108907, "learning_rate": 7.932381891694493e-05, "loss": 0.0012, "step": 4320 }, { "epoch": 2.2285714285714286, "grad_norm": 0.04293091701214276, "learning_rate": 7.920242457356526e-05, "loss": 0.0012, "step": 4330 }, { "epoch": 2.2337194337194335, "grad_norm": 0.04093195000280922, "learning_rate": 7.908076841081895e-05, "loss": 0.0013, "step": 4340 }, { "epoch": 2.2388674388674388, "grad_norm": 0.0424664722931835, "learning_rate": 7.895885151943527e-05, "loss": 0.0012, "step": 4350 }, { "epoch": 2.244015444015444, "grad_norm": 0.0821934383463096, "learning_rate": 7.883667499248111e-05, "loss": 0.0013, "step": 4360 }, { "epoch": 2.2491634491634493, "grad_norm": 0.039405424949199575, "learning_rate": 7.871423992535116e-05, "loss": 0.0012, "step": 4370 }, { "epoch": 2.254311454311454, "grad_norm": 0.04228642502788665, "learning_rate": 7.859154741575809e-05, "loss": 0.0012, "step": 4380 }, { "epoch": 2.2594594594594595, "grad_norm": 0.04986484606810717, "learning_rate": 7.846859856372275e-05, "loss": 0.0012, "step": 4390 }, { "epoch": 2.2646074646074648, "grad_norm": 0.06381252118101394, "learning_rate": 7.834539447156424e-05, "loss": 0.0013, "step": 4400 }, { "epoch": 2.2697554697554696, "grad_norm": 0.0476340369024287, "learning_rate": 7.822193624389005e-05, "loss": 0.0013, "step": 4410 }, { "epoch": 2.274903474903475, "grad_norm": 0.053280169836871914, "learning_rate": 7.809822498758618e-05, "loss": 0.0013, "step": 4420 }, { "epoch": 2.28005148005148, "grad_norm": 0.0471759251214204, "learning_rate": 7.797426181180723e-05, "loss": 0.0011, "step": 4430 }, { "epoch": 2.285199485199485, "grad_norm": 0.05750087294114376, "learning_rate": 7.785004782796635e-05, "loss": 0.0012, "step": 4440 }, { "epoch": 2.2903474903474903, "grad_norm": 0.04433322895287332, "learning_rate": 7.772558414972546e-05, "loss": 0.0012, "step": 4450 }, { "epoch": 2.2954954954954956, "grad_norm": 0.05145684185533547, "learning_rate": 7.760087189298509e-05, "loss": 0.0012, "step": 4460 }, { "epoch": 2.300643500643501, "grad_norm": 0.05085435178874306, "learning_rate": 7.747591217587443e-05, "loss": 0.0012, "step": 4470 }, { "epoch": 2.3057915057915057, "grad_norm": 0.05084038319068861, "learning_rate": 7.735070611874137e-05, "loss": 0.0013, "step": 4480 }, { "epoch": 2.310939510939511, "grad_norm": 0.046795622343555554, "learning_rate": 7.722525484414236e-05, "loss": 0.0013, "step": 4490 }, { "epoch": 2.3160875160875163, "grad_norm": 0.043906153913954984, "learning_rate": 7.70995594768324e-05, "loss": 0.0014, "step": 4500 }, { "epoch": 2.321235521235521, "grad_norm": 0.059773455723487405, "learning_rate": 7.6973621143755e-05, "loss": 0.0013, "step": 4510 }, { "epoch": 2.3263835263835264, "grad_norm": 0.05604359193591503, "learning_rate": 7.684744097403191e-05, "loss": 0.0012, "step": 4520 }, { "epoch": 2.3315315315315317, "grad_norm": 0.04675320396248994, "learning_rate": 7.672102009895321e-05, "loss": 0.0013, "step": 4530 }, { "epoch": 2.3366795366795365, "grad_norm": 0.04509813399138995, "learning_rate": 7.659435965196701e-05, "loss": 0.0013, "step": 4540 }, { "epoch": 2.341827541827542, "grad_norm": 0.0576880617509529, "learning_rate": 7.646746076866939e-05, "loss": 0.0014, "step": 4550 }, { "epoch": 2.346975546975547, "grad_norm": 0.06321893688694276, "learning_rate": 7.634032458679412e-05, "loss": 0.0018, "step": 4560 }, { "epoch": 2.352123552123552, "grad_norm": 0.1633121647089786, "learning_rate": 7.621295224620253e-05, "loss": 0.0035, "step": 4570 }, { "epoch": 2.3572715572715572, "grad_norm": 0.08921315729148695, "learning_rate": 7.608534488887327e-05, "loss": 0.0022, "step": 4580 }, { "epoch": 2.3624195624195625, "grad_norm": 0.057771731238916685, "learning_rate": 7.595750365889209e-05, "loss": 0.0019, "step": 4590 }, { "epoch": 2.3675675675675674, "grad_norm": 0.046773705729556504, "learning_rate": 7.582942970244155e-05, "loss": 0.0016, "step": 4600 }, { "epoch": 2.3727155727155727, "grad_norm": 0.061357062741325005, "learning_rate": 7.570112416779078e-05, "loss": 0.0016, "step": 4610 }, { "epoch": 2.377863577863578, "grad_norm": 0.06208953521085493, "learning_rate": 7.557258820528514e-05, "loss": 0.0015, "step": 4620 }, { "epoch": 2.383011583011583, "grad_norm": 0.05563724202044124, "learning_rate": 7.544382296733592e-05, "loss": 0.0014, "step": 4630 }, { "epoch": 2.388159588159588, "grad_norm": 0.054185493414282304, "learning_rate": 7.531482960841004e-05, "loss": 0.0013, "step": 4640 }, { "epoch": 2.3933075933075934, "grad_norm": 0.04752869142758457, "learning_rate": 7.518560928501969e-05, "loss": 0.0013, "step": 4650 }, { "epoch": 2.3984555984555986, "grad_norm": 0.05632456808738891, "learning_rate": 7.505616315571192e-05, "loss": 0.0014, "step": 4660 }, { "epoch": 2.4036036036036035, "grad_norm": 0.04523988270062936, "learning_rate": 7.492649238105828e-05, "loss": 0.0012, "step": 4670 }, { "epoch": 2.4087516087516088, "grad_norm": 0.04001437246775281, "learning_rate": 7.479659812364448e-05, "loss": 0.0012, "step": 4680 }, { "epoch": 2.413899613899614, "grad_norm": 0.04689498867230884, "learning_rate": 7.466648154805982e-05, "loss": 0.0011, "step": 4690 }, { "epoch": 2.419047619047619, "grad_norm": 0.04711399995411257, "learning_rate": 7.453614382088687e-05, "loss": 0.0012, "step": 4700 }, { "epoch": 2.424195624195624, "grad_norm": 0.050401789341980335, "learning_rate": 7.440558611069095e-05, "loss": 0.0012, "step": 4710 }, { "epoch": 2.4293436293436295, "grad_norm": 0.04507054029023686, "learning_rate": 7.427480958800974e-05, "loss": 0.0013, "step": 4720 }, { "epoch": 2.4344916344916343, "grad_norm": 0.05114129043648352, "learning_rate": 7.414381542534266e-05, "loss": 0.0013, "step": 4730 }, { "epoch": 2.4396396396396396, "grad_norm": 0.050858156890527396, "learning_rate": 7.401260479714041e-05, "loss": 0.0012, "step": 4740 }, { "epoch": 2.444787644787645, "grad_norm": 0.046113564407026235, "learning_rate": 7.388117887979449e-05, "loss": 0.0011, "step": 4750 }, { "epoch": 2.44993564993565, "grad_norm": 0.04014385370106455, "learning_rate": 7.37495388516266e-05, "loss": 0.001, "step": 4760 }, { "epoch": 2.455083655083655, "grad_norm": 0.04803165364348251, "learning_rate": 7.361768589287808e-05, "loss": 0.0011, "step": 4770 }, { "epoch": 2.4602316602316603, "grad_norm": 0.033866842151233145, "learning_rate": 7.348562118569935e-05, "loss": 0.0011, "step": 4780 }, { "epoch": 2.4653796653796656, "grad_norm": 0.051903863558592306, "learning_rate": 7.335334591413925e-05, "loss": 0.0011, "step": 4790 }, { "epoch": 2.4705276705276704, "grad_norm": 0.03814836987756061, "learning_rate": 7.322086126413455e-05, "loss": 0.0011, "step": 4800 }, { "epoch": 2.4756756756756757, "grad_norm": 0.039269243726615485, "learning_rate": 7.308816842349918e-05, "loss": 0.001, "step": 4810 }, { "epoch": 2.480823680823681, "grad_norm": 0.04765524329024219, "learning_rate": 7.295526858191362e-05, "loss": 0.0011, "step": 4820 }, { "epoch": 2.485971685971686, "grad_norm": 0.04951778818784927, "learning_rate": 7.282216293091434e-05, "loss": 0.0012, "step": 4830 }, { "epoch": 2.491119691119691, "grad_norm": 0.04536576824454616, "learning_rate": 7.268885266388296e-05, "loss": 0.0011, "step": 4840 }, { "epoch": 2.4962676962676964, "grad_norm": 0.0515320102284534, "learning_rate": 7.255533897603562e-05, "loss": 0.001, "step": 4850 }, { "epoch": 2.5014157014157012, "grad_norm": 0.04600110428117649, "learning_rate": 7.242162306441231e-05, "loss": 0.001, "step": 4860 }, { "epoch": 2.5065637065637065, "grad_norm": 0.041692098130997376, "learning_rate": 7.228770612786604e-05, "loss": 0.0011, "step": 4870 }, { "epoch": 2.511711711711712, "grad_norm": 0.04488982334958329, "learning_rate": 7.215358936705219e-05, "loss": 0.001, "step": 4880 }, { "epoch": 2.5168597168597167, "grad_norm": 0.048039789785818715, "learning_rate": 7.201927398441765e-05, "loss": 0.0011, "step": 4890 }, { "epoch": 2.522007722007722, "grad_norm": 0.04538093295084992, "learning_rate": 7.18847611841901e-05, "loss": 0.001, "step": 4900 }, { "epoch": 2.5271557271557272, "grad_norm": 0.04709836872218103, "learning_rate": 7.175005217236728e-05, "loss": 0.001, "step": 4910 }, { "epoch": 2.532303732303732, "grad_norm": 0.04505447857446204, "learning_rate": 7.161514815670598e-05, "loss": 0.0011, "step": 4920 }, { "epoch": 2.5374517374517374, "grad_norm": 0.05828384125867241, "learning_rate": 7.148005034671142e-05, "loss": 0.0011, "step": 4930 }, { "epoch": 2.5425997425997426, "grad_norm": 0.046343388951512784, "learning_rate": 7.134475995362623e-05, "loss": 0.001, "step": 4940 }, { "epoch": 2.5477477477477475, "grad_norm": 0.04707163071731421, "learning_rate": 7.120927819041979e-05, "loss": 0.001, "step": 4950 }, { "epoch": 2.5528957528957528, "grad_norm": 0.04306173241714982, "learning_rate": 7.107360627177714e-05, "loss": 0.001, "step": 4960 }, { "epoch": 2.558043758043758, "grad_norm": 0.044851977335899865, "learning_rate": 7.093774541408823e-05, "loss": 0.0011, "step": 4970 }, { "epoch": 2.5631917631917633, "grad_norm": 0.040963399467547255, "learning_rate": 7.080169683543701e-05, "loss": 0.0011, "step": 4980 }, { "epoch": 2.568339768339768, "grad_norm": 0.0524316895531215, "learning_rate": 7.066546175559042e-05, "loss": 0.001, "step": 4990 }, { "epoch": 2.5734877734877735, "grad_norm": 0.04091367424148807, "learning_rate": 7.052904139598752e-05, "loss": 0.001, "step": 5000 }, { "epoch": 2.5786357786357788, "grad_norm": 0.05066184832309622, "learning_rate": 7.039243697972857e-05, "loss": 0.0011, "step": 5010 }, { "epoch": 2.583783783783784, "grad_norm": 0.04236548105650226, "learning_rate": 7.025564973156396e-05, "loss": 0.0009, "step": 5020 }, { "epoch": 2.588931788931789, "grad_norm": 0.04623210387140029, "learning_rate": 7.011868087788333e-05, "loss": 0.001, "step": 5030 }, { "epoch": 2.594079794079794, "grad_norm": 0.03783743894851547, "learning_rate": 6.99815316467045e-05, "loss": 0.0011, "step": 5040 }, { "epoch": 2.5992277992277995, "grad_norm": 0.042867495704172615, "learning_rate": 6.984420326766257e-05, "loss": 0.0011, "step": 5050 }, { "epoch": 2.6043758043758043, "grad_norm": 0.042286428164349944, "learning_rate": 6.970669697199876e-05, "loss": 0.0009, "step": 5060 }, { "epoch": 2.6095238095238096, "grad_norm": 0.06287895185713929, "learning_rate": 6.956901399254943e-05, "loss": 0.001, "step": 5070 }, { "epoch": 2.614671814671815, "grad_norm": 0.055451825887909506, "learning_rate": 6.943115556373502e-05, "loss": 0.001, "step": 5080 }, { "epoch": 2.6198198198198197, "grad_norm": 0.037650966668449, "learning_rate": 6.929312292154908e-05, "loss": 0.0011, "step": 5090 }, { "epoch": 2.624967824967825, "grad_norm": 0.0422236192284774, "learning_rate": 6.915491730354697e-05, "loss": 0.001, "step": 5100 }, { "epoch": 2.6301158301158303, "grad_norm": 0.04148234941603724, "learning_rate": 6.9016539948835e-05, "loss": 0.001, "step": 5110 }, { "epoch": 2.635263835263835, "grad_norm": 0.05471853033140722, "learning_rate": 6.887799209805917e-05, "loss": 0.001, "step": 5120 }, { "epoch": 2.6404118404118404, "grad_norm": 0.04759262188474491, "learning_rate": 6.873927499339407e-05, "loss": 0.0011, "step": 5130 }, { "epoch": 2.6455598455598457, "grad_norm": 0.043132316263661374, "learning_rate": 6.860038987853182e-05, "loss": 0.001, "step": 5140 }, { "epoch": 2.6507078507078505, "grad_norm": 0.045288505100104404, "learning_rate": 6.846133799867082e-05, "loss": 0.0011, "step": 5150 }, { "epoch": 2.655855855855856, "grad_norm": 0.03982061447257109, "learning_rate": 6.832212060050465e-05, "loss": 0.001, "step": 5160 }, { "epoch": 2.661003861003861, "grad_norm": 0.03191140789524671, "learning_rate": 6.818273893221088e-05, "loss": 0.0009, "step": 5170 }, { "epoch": 2.666151866151866, "grad_norm": 0.04219223914359278, "learning_rate": 6.804319424343984e-05, "loss": 0.001, "step": 5180 }, { "epoch": 2.6712998712998712, "grad_norm": 0.04261141040671573, "learning_rate": 6.790348778530347e-05, "loss": 0.001, "step": 5190 }, { "epoch": 2.6764478764478765, "grad_norm": 0.046770608070064175, "learning_rate": 6.77636208103641e-05, "loss": 0.0009, "step": 5200 }, { "epoch": 2.6815958815958814, "grad_norm": 0.05046470198461108, "learning_rate": 6.762359457262318e-05, "loss": 0.0009, "step": 5210 }, { "epoch": 2.6867438867438866, "grad_norm": 0.04578382268098655, "learning_rate": 6.748341032751004e-05, "loss": 0.0011, "step": 5220 }, { "epoch": 2.691891891891892, "grad_norm": 0.03916013043164779, "learning_rate": 6.734306933187068e-05, "loss": 0.001, "step": 5230 }, { "epoch": 2.6970398970398968, "grad_norm": 0.0495431816414551, "learning_rate": 6.720257284395649e-05, "loss": 0.0009, "step": 5240 }, { "epoch": 2.702187902187902, "grad_norm": 0.028887936049044396, "learning_rate": 6.706192212341287e-05, "loss": 0.0008, "step": 5250 }, { "epoch": 2.7073359073359073, "grad_norm": 0.04096988584464123, "learning_rate": 6.692111843126812e-05, "loss": 0.001, "step": 5260 }, { "epoch": 2.7124839124839126, "grad_norm": 0.06842570554905965, "learning_rate": 6.6780163029922e-05, "loss": 0.001, "step": 5270 }, { "epoch": 2.7176319176319175, "grad_norm": 0.0417571647859502, "learning_rate": 6.66390571831344e-05, "loss": 0.001, "step": 5280 }, { "epoch": 2.7227799227799228, "grad_norm": 0.03695050716607467, "learning_rate": 6.649780215601411e-05, "loss": 0.001, "step": 5290 }, { "epoch": 2.727927927927928, "grad_norm": 0.0391813604064568, "learning_rate": 6.635639921500742e-05, "loss": 0.0009, "step": 5300 }, { "epoch": 2.7330759330759333, "grad_norm": 0.04228547410069994, "learning_rate": 6.621484962788671e-05, "loss": 0.0011, "step": 5310 }, { "epoch": 2.738223938223938, "grad_norm": 0.043524209122763294, "learning_rate": 6.607315466373923e-05, "loss": 0.0009, "step": 5320 }, { "epoch": 2.7433719433719435, "grad_norm": 0.05219296497463659, "learning_rate": 6.593131559295551e-05, "loss": 0.0009, "step": 5330 }, { "epoch": 2.7485199485199487, "grad_norm": 0.04065132149131985, "learning_rate": 6.578933368721824e-05, "loss": 0.001, "step": 5340 }, { "epoch": 2.7536679536679536, "grad_norm": 0.040021905728438036, "learning_rate": 6.564721021949061e-05, "loss": 0.001, "step": 5350 }, { "epoch": 2.758815958815959, "grad_norm": 0.03991881038489872, "learning_rate": 6.550494646400506e-05, "loss": 0.0008, "step": 5360 }, { "epoch": 2.763963963963964, "grad_norm": 0.0428200375670312, "learning_rate": 6.536254369625176e-05, "loss": 0.0008, "step": 5370 }, { "epoch": 2.769111969111969, "grad_norm": 0.040946393497408425, "learning_rate": 6.52200031929673e-05, "loss": 0.0008, "step": 5380 }, { "epoch": 2.7742599742599743, "grad_norm": 0.043599458814907815, "learning_rate": 6.507732623212308e-05, "loss": 0.0008, "step": 5390 }, { "epoch": 2.7794079794079796, "grad_norm": 0.04744156312077381, "learning_rate": 6.493451409291394e-05, "loss": 0.0009, "step": 5400 } ], "logging_steps": 10, "max_steps": 11658, "num_input_tokens_seen": 0, "num_train_epochs": 6, "save_steps": 1800, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 484502657826816.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }