{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 2000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.9290919303894043, "learning_rate": 1.8000000000000001e-06, "loss": 0.0256, "step": 10 }, { "grad_norm": 0.8753812909126282, "learning_rate": 3.8000000000000005e-06, "loss": 0.0254, "step": 20 }, { "grad_norm": 0.5441280603408813, "learning_rate": 5.8e-06, "loss": 0.0228, "step": 30 }, { "grad_norm": 0.5367457866668701, "learning_rate": 7.800000000000002e-06, "loss": 0.0246, "step": 40 }, { "grad_norm": 1.0656629800796509, "learning_rate": 9.800000000000001e-06, "loss": 0.0276, "step": 50 }, { "grad_norm": 0.7778517007827759, "learning_rate": 1.18e-05, "loss": 0.033, "step": 60 }, { "grad_norm": 0.5689480304718018, "learning_rate": 1.38e-05, "loss": 0.0257, "step": 70 }, { "grad_norm": 0.43080252408981323, "learning_rate": 1.58e-05, "loss": 0.0289, "step": 80 }, { "grad_norm": 0.8442323803901672, "learning_rate": 1.7800000000000002e-05, "loss": 0.0328, "step": 90 }, { "grad_norm": 0.9511900544166565, "learning_rate": 1.98e-05, "loss": 0.0377, "step": 100 }, { "grad_norm": 0.808724582195282, "learning_rate": 1.9998892765646024e-05, "loss": 0.0329, "step": 110 }, { "grad_norm": 0.5327794551849365, "learning_rate": 1.9995065603657317e-05, "loss": 0.0454, "step": 120 }, { "grad_norm": 0.6913271546363831, "learning_rate": 1.998850589053268e-05, "loss": 0.0275, "step": 130 }, { "grad_norm": 0.46081170439720154, "learning_rate": 1.997921541963218e-05, "loss": 0.0274, "step": 140 }, { "grad_norm": 0.5543823838233948, "learning_rate": 1.9967196730877805e-05, "loss": 0.0357, "step": 150 }, { "grad_norm": 0.3269347548484802, "learning_rate": 1.9952453110059044e-05, "loss": 0.0326, "step": 160 }, { "grad_norm": 0.5915572643280029, "learning_rate": 1.9934988587934624e-05, "loss": 0.0392, "step": 170 }, { "grad_norm": 0.4258320927619934, "learning_rate": 1.99148079391305e-05, "loss": 0.0315, "step": 180 }, { "grad_norm": 0.49777689576148987, "learning_rate": 1.9891916680834565e-05, "loss": 0.0472, "step": 190 }, { "grad_norm": 0.6657916903495789, "learning_rate": 1.9866321071288262e-05, "loss": 0.0426, "step": 200 }, { "grad_norm": 0.6278263926506042, "learning_rate": 1.9838028108075673e-05, "loss": 0.0431, "step": 210 }, { "grad_norm": 0.5320912003517151, "learning_rate": 1.9807045526210435e-05, "loss": 0.0411, "step": 220 }, { "grad_norm": 0.49899357557296753, "learning_rate": 1.9773381796021073e-05, "loss": 0.0412, "step": 230 }, { "grad_norm": 0.4521670639514923, "learning_rate": 1.973704612083529e-05, "loss": 0.0328, "step": 240 }, { "grad_norm": 0.504338800907135, "learning_rate": 1.969804843446387e-05, "loss": 0.0526, "step": 250 }, { "grad_norm": 0.5938769578933716, "learning_rate": 1.965639939848488e-05, "loss": 0.0413, "step": 260 }, { "grad_norm": 0.3719467520713806, "learning_rate": 1.961211039932889e-05, "loss": 0.0508, "step": 270 }, { "grad_norm": 0.4052974283695221, "learning_rate": 1.9565193545166052e-05, "loss": 0.0436, "step": 280 }, { "grad_norm": 0.5165993571281433, "learning_rate": 1.951566166259583e-05, "loss": 0.0368, "step": 290 }, { "grad_norm": 0.4715283513069153, "learning_rate": 1.9463528293140347e-05, "loss": 0.0395, "step": 300 }, { "grad_norm": 0.5041248202323914, "learning_rate": 1.9408807689542257e-05, "loss": 0.0407, "step": 310 }, { "grad_norm": 0.6388036608695984, "learning_rate": 1.935151481186821e-05, "loss": 0.0439, "step": 320 }, { "grad_norm": 0.6267423629760742, "learning_rate": 1.9291665323418888e-05, "loss": 0.0485, "step": 330 }, { "grad_norm": 0.5609102845191956, "learning_rate": 1.9229275586446853e-05, "loss": 0.0416, "step": 340 }, { "grad_norm": 0.485109806060791, "learning_rate": 1.9164362657683223e-05, "loss": 0.0514, "step": 350 }, { "grad_norm": 0.5206034779548645, "learning_rate": 1.9096944283674574e-05, "loss": 0.0483, "step": 360 }, { "grad_norm": 0.5093031525611877, "learning_rate": 1.902703889593119e-05, "loss": 0.0456, "step": 370 }, { "grad_norm": 0.573880672454834, "learning_rate": 1.8954665605888086e-05, "loss": 0.0429, "step": 380 }, { "grad_norm": 1.1320035457611084, "learning_rate": 1.887984419968011e-05, "loss": 0.0573, "step": 390 }, { "grad_norm": 0.410280704498291, "learning_rate": 1.8802595132732636e-05, "loss": 0.0392, "step": 400 }, { "grad_norm": 0.7732188105583191, "learning_rate": 1.8722939524169243e-05, "loss": 0.0513, "step": 410 }, { "grad_norm": 0.5694898366928101, "learning_rate": 1.8640899151037945e-05, "loss": 0.0507, "step": 420 }, { "grad_norm": 0.675169825553894, "learning_rate": 1.8556496442357597e-05, "loss": 0.0534, "step": 430 }, { "grad_norm": 0.6167284846305847, "learning_rate": 1.8469754472985994e-05, "loss": 0.0507, "step": 440 }, { "grad_norm": 0.6989724040031433, "learning_rate": 1.8380696957311447e-05, "loss": 0.0513, "step": 450 }, { "grad_norm": 0.41459566354751587, "learning_rate": 1.8289348242769515e-05, "loss": 0.042, "step": 460 }, { "grad_norm": 0.6491379141807556, "learning_rate": 1.8195733303186634e-05, "loss": 0.0498, "step": 470 }, { "grad_norm": 0.6334497332572937, "learning_rate": 1.809987773195255e-05, "loss": 0.0376, "step": 480 }, { "grad_norm": 0.3463200628757477, "learning_rate": 1.800180773502333e-05, "loss": 0.0549, "step": 490 }, { "grad_norm": 0.6881201863288879, "learning_rate": 1.7901550123756906e-05, "loss": 0.0546, "step": 500 }, { "grad_norm": 0.6690155267715454, "learning_rate": 1.7799132307583133e-05, "loss": 0.0553, "step": 510 }, { "grad_norm": 0.7641336917877197, "learning_rate": 1.769458228651032e-05, "loss": 0.0507, "step": 520 }, { "grad_norm": 0.454863041639328, "learning_rate": 1.75879286434703e-05, "loss": 0.0577, "step": 530 }, { "grad_norm": 0.6699721813201904, "learning_rate": 1.7479200536504154e-05, "loss": 0.0598, "step": 540 }, { "grad_norm": 0.6386680006980896, "learning_rate": 1.7368427690790678e-05, "loss": 0.049, "step": 550 }, { "grad_norm": 0.7113078832626343, "learning_rate": 1.7255640390519838e-05, "loss": 0.0651, "step": 560 }, { "grad_norm": 0.8806372880935669, "learning_rate": 1.7140869470613344e-05, "loss": 0.0645, "step": 570 }, { "grad_norm": 0.7006818056106567, "learning_rate": 1.7024146308294725e-05, "loss": 0.0576, "step": 580 }, { "grad_norm": 0.44408583641052246, "learning_rate": 1.6905502814511085e-05, "loss": 0.0496, "step": 590 }, { "grad_norm": 0.8034731149673462, "learning_rate": 1.6784971425208967e-05, "loss": 0.0611, "step": 600 }, { "grad_norm": 0.7743136286735535, "learning_rate": 1.6662585092466722e-05, "loss": 0.0611, "step": 610 }, { "grad_norm": 0.8685964345932007, "learning_rate": 1.6538377275485694e-05, "loss": 0.0612, "step": 620 }, { "grad_norm": 0.5246061086654663, "learning_rate": 1.6412381931442837e-05, "loss": 0.0511, "step": 630 }, { "grad_norm": 0.6122261881828308, "learning_rate": 1.628463350620716e-05, "loss": 0.0542, "step": 640 }, { "grad_norm": 0.5953839421272278, "learning_rate": 1.6155166924922564e-05, "loss": 0.0513, "step": 650 }, { "grad_norm": 0.7195720672607422, "learning_rate": 1.6024017582459652e-05, "loss": 0.0537, "step": 660 }, { "grad_norm": 0.5591540932655334, "learning_rate": 1.589122133373914e-05, "loss": 0.0397, "step": 670 }, { "grad_norm": 0.5585523247718811, "learning_rate": 1.575681448392946e-05, "loss": 0.0623, "step": 680 }, { "grad_norm": 0.5158264636993408, "learning_rate": 1.5620833778521306e-05, "loss": 0.0457, "step": 690 }, { "grad_norm": 0.8843851685523987, "learning_rate": 1.5483316393281783e-05, "loss": 0.0479, "step": 700 }, { "grad_norm": 0.8452252745628357, "learning_rate": 1.5344299924090915e-05, "loss": 0.0572, "step": 710 }, { "grad_norm": 0.45865270495414734, "learning_rate": 1.52038223766633e-05, "loss": 0.0421, "step": 720 }, { "grad_norm": 0.7401474714279175, "learning_rate": 1.5061922156157746e-05, "loss": 0.0502, "step": 730 }, { "grad_norm": 0.5065820813179016, "learning_rate": 1.4918638056677682e-05, "loss": 0.0381, "step": 740 }, { "grad_norm": 0.8904169797897339, "learning_rate": 1.4774009250665215e-05, "loss": 0.0429, "step": 750 }, { "grad_norm": 0.5113430619239807, "learning_rate": 1.4628075278191803e-05, "loss": 0.0352, "step": 760 }, { "grad_norm": 0.6518237590789795, "learning_rate": 1.4480876036148376e-05, "loss": 0.042, "step": 770 }, { "grad_norm": 0.5012675523757935, "learning_rate": 1.4332451767337938e-05, "loss": 0.0381, "step": 780 }, { "grad_norm": 0.6119031310081482, "learning_rate": 1.418284304947357e-05, "loss": 0.0387, "step": 790 }, { "grad_norm": 0.3946932554244995, "learning_rate": 1.4032090784084904e-05, "loss": 0.0344, "step": 800 }, { "grad_norm": 0.5871202945709229, "learning_rate": 1.3880236185336045e-05, "loss": 0.0334, "step": 810 }, { "grad_norm": 0.7659527063369751, "learning_rate": 1.3727320768758034e-05, "loss": 0.0524, "step": 820 }, { "grad_norm": 0.6022297739982605, "learning_rate": 1.357338633989891e-05, "loss": 0.0483, "step": 830 }, { "grad_norm": 0.4584524631500244, "learning_rate": 1.3418474982894499e-05, "loss": 0.0305, "step": 840 }, { "grad_norm": 0.6703428030014038, "learning_rate": 1.3262629048963028e-05, "loss": 0.0425, "step": 850 }, { "grad_norm": 0.5280142426490784, "learning_rate": 1.3105891144826715e-05, "loss": 0.0391, "step": 860 }, { "grad_norm": 0.7856833934783936, "learning_rate": 1.2948304121063538e-05, "loss": 0.0417, "step": 870 }, { "grad_norm": 0.4869779348373413, "learning_rate": 1.2789911060392295e-05, "loss": 0.0309, "step": 880 }, { "grad_norm": 0.48766058683395386, "learning_rate": 1.263075526589423e-05, "loss": 0.0316, "step": 890 }, { "grad_norm": 0.3586791157722473, "learning_rate": 1.2470880249174397e-05, "loss": 0.0251, "step": 900 }, { "grad_norm": 0.6049955487251282, "learning_rate": 1.2310329718466023e-05, "loss": 0.0283, "step": 910 }, { "grad_norm": 0.4073525369167328, "learning_rate": 1.2149147566681126e-05, "loss": 0.0325, "step": 920 }, { "grad_norm": 0.876194417476654, "learning_rate": 1.1987377859410598e-05, "loss": 0.0422, "step": 930 }, { "grad_norm": 0.5535964965820312, "learning_rate": 1.1825064822877153e-05, "loss": 0.0263, "step": 940 }, { "grad_norm": 0.28055599331855774, "learning_rate": 1.1662252831844297e-05, "loss": 0.0304, "step": 950 }, { "grad_norm": 0.41647839546203613, "learning_rate": 1.14989863974847e-05, "loss": 0.0358, "step": 960 }, { "grad_norm": 0.4186880588531494, "learning_rate": 1.1335310155211319e-05, "loss": 0.0271, "step": 970 }, { "grad_norm": 0.6154966354370117, "learning_rate": 1.1171268852474492e-05, "loss": 0.0284, "step": 980 }, { "grad_norm": 0.6328971982002258, "learning_rate": 1.1006907336528445e-05, "loss": 0.0274, "step": 990 }, { "grad_norm": 0.35281991958618164, "learning_rate": 1.0842270542170498e-05, "loss": 0.0239, "step": 1000 }, { "grad_norm": 0.3426606357097626, "learning_rate": 1.067740347945634e-05, "loss": 0.035, "step": 1010 }, { "grad_norm": 0.3788098990917206, "learning_rate": 1.0512351221394731e-05, "loss": 0.0263, "step": 1020 }, { "grad_norm": 0.8431473970413208, "learning_rate": 1.0347158891624992e-05, "loss": 0.031, "step": 1030 }, { "grad_norm": 0.6758059859275818, "learning_rate": 1.0181871652080637e-05, "loss": 0.0441, "step": 1040 }, { "grad_norm": 0.5647501945495605, "learning_rate": 1.0016534690642584e-05, "loss": 0.0435, "step": 1050 }, { "grad_norm": 0.5106040835380554, "learning_rate": 9.851193208785206e-06, "loss": 0.0561, "step": 1060 }, { "grad_norm": 0.40076860785484314, "learning_rate": 9.685892409218718e-06, "loss": 0.0416, "step": 1070 }, { "grad_norm": 0.586910605430603, "learning_rate": 9.520677483531205e-06, "loss": 0.0495, "step": 1080 }, { "grad_norm": 0.6157604455947876, "learning_rate": 9.35559359983369e-06, "loss": 0.0454, "step": 1090 }, { "grad_norm": 0.4568628668785095, "learning_rate": 9.190685890411628e-06, "loss": 0.041, "step": 1100 }, { "grad_norm": 0.5501033067703247, "learning_rate": 9.02599943938617e-06, "loss": 0.0421, "step": 1110 }, { "grad_norm": 0.5676171779632568, "learning_rate": 8.861579270388649e-06, "loss": 0.037, "step": 1120 }, { "grad_norm": 0.4058741629123688, "learning_rate": 8.697470334251543e-06, "loss": 0.0464, "step": 1130 }, { "grad_norm": 0.6203457713127136, "learning_rate": 8.533717496719374e-06, "loss": 0.0368, "step": 1140 }, { "grad_norm": 0.44810524582862854, "learning_rate": 8.370365526182862e-06, "loss": 0.0492, "step": 1150 }, { "grad_norm": 0.51730877161026, "learning_rate": 8.207459081439696e-06, "loss": 0.0399, "step": 1160 }, { "grad_norm": 0.6902947425842285, "learning_rate": 8.045042699485256e-06, "loss": 0.048, "step": 1170 }, { "grad_norm": 0.514245867729187, "learning_rate": 7.883160783336646e-06, "loss": 0.0458, "step": 1180 }, { "grad_norm": 0.6786456108093262, "learning_rate": 7.721857589893364e-06, "loss": 0.0494, "step": 1190 }, { "grad_norm": 0.5476706027984619, "learning_rate": 7.561177217837894e-06, "loss": 0.0456, "step": 1200 }, { "grad_norm": 0.7459077835083008, "learning_rate": 7.401163595579573e-06, "loss": 0.0526, "step": 1210 }, { "grad_norm": 0.5312514901161194, "learning_rate": 7.2418604692450015e-06, "loss": 0.0496, "step": 1220 }, { "grad_norm": 0.49069681763648987, "learning_rate": 7.083311390718285e-06, "loss": 0.045, "step": 1230 }, { "grad_norm": 0.6538358926773071, "learning_rate": 6.925559705734394e-06, "loss": 0.0519, "step": 1240 }, { "grad_norm": 0.6089681386947632, "learning_rate": 6.7686485420288576e-06, "loss": 0.0423, "step": 1250 }, { "grad_norm": 0.47313421964645386, "learning_rate": 6.612620797547087e-06, "loss": 0.0352, "step": 1260 }, { "grad_norm": 0.30322036147117615, "learning_rate": 6.457519128716496e-06, "loss": 0.0348, "step": 1270 }, { "grad_norm": 0.34919723868370056, "learning_rate": 6.303385938784663e-06, "loss": 0.0386, "step": 1280 }, { "grad_norm": 0.3409254550933838, "learning_rate": 6.150263366226703e-06, "loss": 0.0336, "step": 1290 }, { "grad_norm": 0.3261924088001251, "learning_rate": 5.9981932732250356e-06, "loss": 0.0304, "step": 1300 }, { "grad_norm": 0.3297593891620636, "learning_rate": 5.8472172342246816e-06, "loss": 0.0334, "step": 1310 }, { "grad_norm": 0.28802627325057983, "learning_rate": 5.697376524567205e-06, "loss": 0.0358, "step": 1320 }, { "grad_norm": 0.4600858986377716, "learning_rate": 5.548712109206487e-06, "loss": 0.0383, "step": 1330 }, { "grad_norm": 0.621453046798706, "learning_rate": 5.401264631509277e-06, "loss": 0.0396, "step": 1340 }, { "grad_norm": 0.5540786385536194, "learning_rate": 5.255074402143727e-06, "loss": 0.0312, "step": 1350 }, { "grad_norm": 0.4618375897407532, "learning_rate": 5.110181388058842e-06, "loss": 0.0401, "step": 1360 }, { "grad_norm": 0.4480360448360443, "learning_rate": 4.966625201557931e-06, "loss": 0.0354, "step": 1370 }, { "grad_norm": 0.5192937254905701, "learning_rate": 4.824445089468975e-06, "loss": 0.0298, "step": 1380 }, { "grad_norm": 0.4659355580806732, "learning_rate": 4.683679922414964e-06, "loss": 0.0391, "step": 1390 }, { "grad_norm": 0.4526372253894806, "learning_rate": 4.544368184187039e-06, "loss": 0.03, "step": 1400 }, { "grad_norm": 0.3998863995075226, "learning_rate": 4.406547961223412e-06, "loss": 0.0462, "step": 1410 }, { "grad_norm": 0.5105950832366943, "learning_rate": 4.2702569321969145e-06, "loss": 0.0326, "step": 1420 }, { "grad_norm": 0.7159051895141602, "learning_rate": 4.135532357714026e-06, "loss": 0.0441, "step": 1430 }, { "grad_norm": 0.46011897921562195, "learning_rate": 4.002411070128197e-06, "loss": 0.0489, "step": 1440 }, { "grad_norm": 0.37638065218925476, "learning_rate": 3.8709294634702374e-06, "loss": 0.0576, "step": 1450 }, { "grad_norm": 0.5009196996688843, "learning_rate": 3.7411234834985765e-06, "loss": 0.0524, "step": 1460 }, { "grad_norm": 0.40672561526298523, "learning_rate": 3.6130286178720343e-06, "loss": 0.0403, "step": 1470 }, { "grad_norm": 0.38205215334892273, "learning_rate": 3.4866798864478523e-06, "loss": 0.0485, "step": 1480 }, { "grad_norm": 0.41737431287765503, "learning_rate": 3.3621118317076183e-06, "loss": 0.0475, "step": 1490 }, { "grad_norm": 0.6038498282432556, "learning_rate": 3.239358509313695e-06, "loss": 0.054, "step": 1500 }, { "grad_norm": 0.2340111881494522, "learning_rate": 3.118453478798743e-06, "loss": 0.0404, "step": 1510 }, { "grad_norm": 0.3504510223865509, "learning_rate": 2.999429794390869e-06, "loss": 0.0468, "step": 1520 }, { "grad_norm": 0.2890276610851288, "learning_rate": 2.882319995976949e-06, "loss": 0.0409, "step": 1530 }, { "grad_norm": 0.7779890298843384, "learning_rate": 2.7671561002065362e-06, "loss": 0.0399, "step": 1540 }, { "grad_norm": 0.48023533821105957, "learning_rate": 2.65396959173883e-06, "loss": 0.0507, "step": 1550 }, { "grad_norm": 0.5299713611602783, "learning_rate": 2.542791414635085e-06, "loss": 0.0347, "step": 1560 }, { "grad_norm": 0.4216732382774353, "learning_rate": 2.4336519638988134e-06, "loss": 0.0381, "step": 1570 }, { "grad_norm": 0.4404115378856659, "learning_rate": 2.326581077166097e-06, "loss": 0.0485, "step": 1580 }, { "grad_norm": 0.3582431375980377, "learning_rate": 2.221608026548271e-06, "loss": 0.044, "step": 1590 }, { "grad_norm": 0.9148777723312378, "learning_rate": 2.118761510629226e-06, "loss": 0.0528, "step": 1600 }, { "grad_norm": 0.7486399412155151, "learning_rate": 2.0180696466194983e-06, "loss": 0.0432, "step": 1610 }, { "grad_norm": 0.3516605496406555, "learning_rate": 1.919559962669305e-06, "loss": 0.0455, "step": 1620 }, { "grad_norm": 0.4082990288734436, "learning_rate": 1.8232593903426266e-06, "loss": 0.0481, "step": 1630 }, { "grad_norm": 0.5837849974632263, "learning_rate": 1.7291942572543806e-06, "loss": 0.0473, "step": 1640 }, { "grad_norm": 0.4245234429836273, "learning_rate": 1.6373902798727226e-06, "loss": 0.0359, "step": 1650 }, { "grad_norm": 0.365691602230072, "learning_rate": 1.5478725564884045e-06, "loss": 0.0526, "step": 1660 }, { "grad_norm": 0.35451436042785645, "learning_rate": 1.4606655603531762e-06, "loss": 0.0334, "step": 1670 }, { "grad_norm": 0.5978314280509949, "learning_rate": 1.3757931329890218e-06, "loss": 0.0349, "step": 1680 }, { "grad_norm": 0.3389391303062439, "learning_rate": 1.293278477670139e-06, "loss": 0.0478, "step": 1690 }, { "grad_norm": 0.3349228799343109, "learning_rate": 1.2131441530793863e-06, "loss": 0.0294, "step": 1700 }, { "grad_norm": 0.41341710090637207, "learning_rate": 1.135412067140974e-06, "loss": 0.0345, "step": 1710 }, { "grad_norm": 0.5619969367980957, "learning_rate": 1.0601034710310431e-06, "loss": 0.0328, "step": 1720 }, { "grad_norm": 0.39567798376083374, "learning_rate": 9.872389533678207e-07, "loss": 0.0289, "step": 1730 }, { "grad_norm": 0.5375956296920776, "learning_rate": 9.168384345828885e-07, "loss": 0.0309, "step": 1740 }, { "grad_norm": 0.2654189467430115, "learning_rate": 8.489211614751359e-07, "loss": 0.0273, "step": 1750 }, { "grad_norm": 0.369599312543869, "learning_rate": 7.835057019488823e-07, "loss": 0.0228, "step": 1760 }, { "grad_norm": 0.37427300214767456, "learning_rate": 7.206099399375921e-07, "loss": 0.0273, "step": 1770 }, { "grad_norm": 0.21921221911907196, "learning_rate": 6.602510705145893e-07, "loss": 0.0239, "step": 1780 }, { "grad_norm": 0.29054954648017883, "learning_rate": 6.024455951920949e-07, "loss": 0.031, "step": 1790 }, { "grad_norm": 0.2911668121814728, "learning_rate": 5.47209317409878e-07, "loss": 0.0306, "step": 1800 }, { "grad_norm": 0.5516527891159058, "learning_rate": 4.945573382147517e-07, "loss": 0.0285, "step": 1810 }, { "grad_norm": 0.31199005246162415, "learning_rate": 4.4450405213210424e-07, "loss": 0.0224, "step": 1820 }, { "grad_norm": 0.338760644197464, "learning_rate": 3.9706314323056936e-07, "loss": 0.0259, "step": 1830 }, { "grad_norm": 0.4687983989715576, "learning_rate": 3.522475813809467e-07, "loss": 0.0326, "step": 1840 }, { "grad_norm": 0.2302476465702057, "learning_rate": 3.100696187103658e-07, "loss": 0.0329, "step": 1850 }, { "grad_norm": 0.2902008593082428, "learning_rate": 2.705407862526765e-07, "loss": 0.0294, "step": 1860 }, { "grad_norm": 0.2698631286621094, "learning_rate": 2.3367189079597785e-07, "loss": 0.0278, "step": 1870 }, { "grad_norm": 0.20767030119895935, "learning_rate": 1.9947301192815738e-07, "loss": 0.0326, "step": 1880 }, { "grad_norm": 0.23216848075389862, "learning_rate": 1.679534992812215e-07, "loss": 0.0327, "step": 1890 }, { "grad_norm": 0.3182239234447479, "learning_rate": 1.3912196997520776e-07, "loss": 0.0357, "step": 1900 }, { "grad_norm": 0.4465329945087433, "learning_rate": 1.129863062623482e-07, "loss": 0.0288, "step": 1910 }, { "grad_norm": 0.24376554787158966, "learning_rate": 8.955365337213861e-08, "loss": 0.0173, "step": 1920 }, { "grad_norm": 0.31865909695625305, "learning_rate": 6.883041755790576e-08, "loss": 0.0244, "step": 1930 }, { "grad_norm": 0.25592926144599915, "learning_rate": 5.0822264345400384e-08, "loss": 0.0225, "step": 1940 }, { "grad_norm": 0.2635904848575592, "learning_rate": 3.553411698390075e-08, "loss": 0.0229, "step": 1950 }, { "grad_norm": 0.7655121088027954, "learning_rate": 2.2970155100244807e-08, "loss": 0.0275, "step": 1960 }, { "grad_norm": 0.27914825081825256, "learning_rate": 1.3133813556159835e-08, "loss": 0.0233, "step": 1970 }, { "grad_norm": 0.36059820652008057, "learning_rate": 6.02778150921024e-09, "loss": 0.034, "step": 1980 }, { "grad_norm": 0.2451169192790985, "learning_rate": 1.6540016776045442e-09, "loss": 0.0273, "step": 1990 }, { "grad_norm": 0.48328912258148193, "learning_rate": 1.366980907580029e-11, "loss": 0.0335, "step": 2000 } ], "logging_steps": 10, "max_steps": 2000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }