{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8780594885303479, "global_step": 2000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 4.999914338886042e-05, "loss": 1.2139, "step": 10 }, { "epoch": 0.01, "learning_rate": 4.999390875559143e-05, "loss": 0.4338, "step": 20 }, { "epoch": 0.01, "learning_rate": 4.99839163793675e-05, "loss": 0.1646, "step": 30 }, { "epoch": 0.02, "learning_rate": 4.997085692392521e-05, "loss": 0.1648, "step": 40 }, { "epoch": 0.02, "learning_rate": 4.995183082464269e-05, "loss": 0.081, "step": 50 }, { "epoch": 0.03, "learning_rate": 4.9928054992195985e-05, "loss": 0.0475, "step": 60 }, { "epoch": 0.03, "learning_rate": 4.98995339524598e-05, "loss": 0.0529, "step": 70 }, { "epoch": 0.04, "learning_rate": 4.986981233301252e-05, "loss": 0.1633, "step": 80 }, { "epoch": 0.04, "learning_rate": 4.983229110247534e-05, "loss": 0.0649, "step": 90 }, { "epoch": 0.04, "learning_rate": 4.97900428938891e-05, "loss": 0.0791, "step": 100 }, { "epoch": 0.05, "learning_rate": 4.974798456824312e-05, "loss": 0.1031, "step": 110 }, { "epoch": 0.05, "learning_rate": 4.9696778000097424e-05, "loss": 0.0449, "step": 120 }, { "epoch": 0.06, "learning_rate": 4.964087024964812e-05, "loss": 0.0279, "step": 130 }, { "epoch": 0.06, "learning_rate": 4.9580271959276445e-05, "loss": 0.041, "step": 140 }, { "epoch": 0.07, "learning_rate": 4.9514994664236535e-05, "loss": 0.0454, "step": 150 }, { "epoch": 0.07, "learning_rate": 4.944505079045958e-05, "loss": 0.016, "step": 160 }, { "epoch": 0.07, "learning_rate": 4.9378122365147536e-05, "loss": 0.1362, "step": 170 }, { "epoch": 0.08, "learning_rate": 4.929934940465946e-05, "loss": 0.0186, "step": 180 }, { "epoch": 0.08, "learning_rate": 4.921595091482461e-05, "loss": 0.0381, "step": 190 }, { "epoch": 0.09, "learning_rate": 4.913695054926396e-05, "loss": 0.0952, "step": 200 }, { "epoch": 0.09, "learning_rate": 4.904480801572673e-05, "loss": 0.0432, "step": 210 }, { "epoch": 0.1, "learning_rate": 4.8948088406348404e-05, "loss": 0.0343, "step": 220 }, { "epoch": 0.1, "learning_rate": 4.884681013229645e-05, "loss": 0.1469, "step": 230 }, { "epoch": 0.11, "learning_rate": 4.874099247250798e-05, "loss": 0.0432, "step": 240 }, { "epoch": 0.11, "learning_rate": 4.8641892034320465e-05, "loss": 0.138, "step": 250 }, { "epoch": 0.11, "learning_rate": 4.85275057467904e-05, "loss": 0.0637, "step": 260 }, { "epoch": 0.12, "learning_rate": 4.8408640855060846e-05, "loss": 0.0268, "step": 270 }, { "epoch": 0.12, "learning_rate": 4.828531998578885e-05, "loss": 0.0583, "step": 280 }, { "epoch": 0.13, "learning_rate": 4.815756661385392e-05, "loss": 0.0248, "step": 290 }, { "epoch": 0.13, "learning_rate": 4.802540505788951e-05, "loss": 0.0247, "step": 300 }, { "epoch": 0.14, "learning_rate": 4.788886047565377e-05, "loss": 0.074, "step": 310 }, { "epoch": 0.14, "learning_rate": 4.7747958859240626e-05, "loss": 0.0232, "step": 320 }, { "epoch": 0.14, "learning_rate": 4.760272703013206e-05, "loss": 0.0195, "step": 330 }, { "epoch": 0.15, "learning_rate": 4.74531926340924e-05, "loss": 0.0382, "step": 340 }, { "epoch": 0.15, "learning_rate": 4.729938413590588e-05, "loss": 0.037, "step": 350 }, { "epoch": 0.16, "learning_rate": 4.714133081395811e-05, "loss": 0.0196, "step": 360 }, { "epoch": 0.16, "learning_rate": 4.6995478349706145e-05, "loss": 0.0784, "step": 370 }, { "epoch": 0.17, "learning_rate": 4.68294434139043e-05, "loss": 0.0356, "step": 380 }, { "epoch": 0.17, "learning_rate": 4.665925311042354e-05, "loss": 0.0135, "step": 390 }, { "epoch": 0.18, "learning_rate": 4.648493983602601e-05, "loss": 0.0417, "step": 400 }, { "epoch": 0.18, "learning_rate": 4.630653677230654e-05, "loss": 0.0064, "step": 410 }, { "epoch": 0.18, "learning_rate": 4.614250529713827e-05, "loss": 0.1264, "step": 420 }, { "epoch": 0.19, "learning_rate": 4.595642583205011e-05, "loss": 0.0072, "step": 430 }, { "epoch": 0.19, "learning_rate": 4.5766357183503764e-05, "loss": 0.0022, "step": 440 }, { "epoch": 0.2, "learning_rate": 4.559191453574582e-05, "loss": 0.0002, "step": 450 }, { "epoch": 0.2, "learning_rate": 4.539436673859789e-05, "loss": 0.1255, "step": 460 }, { "epoch": 0.21, "learning_rate": 4.519293674936788e-05, "loss": 0.0577, "step": 470 }, { "epoch": 0.21, "learning_rate": 4.498766291148263e-05, "loss": 0.0095, "step": 480 }, { "epoch": 0.22, "learning_rate": 4.479966224737879e-05, "loss": 0.0778, "step": 490 }, { "epoch": 0.22, "learning_rate": 4.458719334778153e-05, "loss": 0.0134, "step": 500 }, { "epoch": 0.22, "learning_rate": 4.4370995906479616e-05, "loss": 0.0127, "step": 510 }, { "epoch": 0.23, "learning_rate": 4.415111107797445e-05, "loss": 0.0194, "step": 520 }, { "epoch": 0.23, "learning_rate": 4.3927580718684106e-05, "loss": 0.0121, "step": 530 }, { "epoch": 0.24, "learning_rate": 4.372332162017128e-05, "loss": 0.1112, "step": 540 }, { "epoch": 0.24, "learning_rate": 4.3492982545674274e-05, "loss": 0.0298, "step": 550 }, { "epoch": 0.25, "learning_rate": 4.325912321917175e-05, "loss": 0.0719, "step": 560 }, { "epoch": 0.25, "learning_rate": 4.302178815721267e-05, "loss": 0.0508, "step": 570 }, { "epoch": 0.25, "learning_rate": 4.278102253797347e-05, "loss": 0.0038, "step": 580 }, { "epoch": 0.26, "learning_rate": 4.2536872192658036e-05, "loss": 0.0831, "step": 590 }, { "epoch": 0.26, "learning_rate": 4.228938359677354e-05, "loss": 0.0713, "step": 600 }, { "epoch": 0.27, "learning_rate": 4.20386038612835e-05, "loss": 0.0755, "step": 610 }, { "epoch": 0.27, "learning_rate": 4.1784580723639923e-05, "loss": 0.0503, "step": 620 }, { "epoch": 0.28, "learning_rate": 4.1527362538696204e-05, "loss": 0.0098, "step": 630 }, { "epoch": 0.28, "learning_rate": 4.1266998269502456e-05, "loss": 0.0212, "step": 640 }, { "epoch": 0.29, "learning_rate": 4.1003537477985154e-05, "loss": 0.078, "step": 650 }, { "epoch": 0.29, "learning_rate": 4.073703031551264e-05, "loss": 0.0195, "step": 660 }, { "epoch": 0.29, "learning_rate": 4.046752751334857e-05, "loss": 0.1526, "step": 670 }, { "epoch": 0.3, "learning_rate": 4.0195080372994864e-05, "loss": 0.001, "step": 680 }, { "epoch": 0.3, "learning_rate": 3.991974075642621e-05, "loss": 0.0041, "step": 690 }, { "epoch": 0.31, "learning_rate": 3.96695053421277e-05, "loss": 0.1016, "step": 700 }, { "epoch": 0.31, "learning_rate": 3.938881486442823e-05, "loss": 0.0073, "step": 710 }, { "epoch": 0.32, "learning_rate": 3.9105385388070305e-05, "loss": 0.0134, "step": 720 }, { "epoch": 0.32, "learning_rate": 3.8819270865583214e-05, "loss": 0.093, "step": 730 }, { "epoch": 0.32, "learning_rate": 3.853052576061115e-05, "loss": 0.0022, "step": 740 }, { "epoch": 0.33, "learning_rate": 3.823920503754575e-05, "loss": 0.0115, "step": 750 }, { "epoch": 0.33, "learning_rate": 3.7945364151063236e-05, "loss": 0.0181, "step": 760 }, { "epoch": 0.34, "learning_rate": 3.764905903556832e-05, "loss": 0.0206, "step": 770 }, { "epoch": 0.34, "learning_rate": 3.7350346094546725e-05, "loss": 0.0263, "step": 780 }, { "epoch": 0.35, "learning_rate": 3.704928218982845e-05, "loss": 0.0438, "step": 790 }, { "epoch": 0.35, "learning_rate": 3.674592463076377e-05, "loss": 0.018, "step": 800 }, { "epoch": 0.36, "learning_rate": 3.644033116331406e-05, "loss": 0.0764, "step": 810 }, { "epoch": 0.36, "learning_rate": 3.613255995905948e-05, "loss": 0.0365, "step": 820 }, { "epoch": 0.36, "learning_rate": 3.5822669604125684e-05, "loss": 0.0045, "step": 830 }, { "epoch": 0.37, "learning_rate": 3.5510719088031576e-05, "loss": 0.0825, "step": 840 }, { "epoch": 0.37, "learning_rate": 3.5196767792460286e-05, "loss": 0.0429, "step": 850 }, { "epoch": 0.38, "learning_rate": 3.4912550346049136e-05, "loss": 0.0981, "step": 860 }, { "epoch": 0.38, "learning_rate": 3.4594962520894233e-05, "loss": 0.0269, "step": 870 }, { "epoch": 0.39, "learning_rate": 3.4275548236109376e-05, "loss": 0.0186, "step": 880 }, { "epoch": 0.39, "learning_rate": 3.39543682941516e-05, "loss": 0.0555, "step": 890 }, { "epoch": 0.4, "learning_rate": 3.363148383358155e-05, "loss": 0.0309, "step": 900 }, { "epoch": 0.4, "learning_rate": 3.330695631742537e-05, "loss": 0.0541, "step": 910 }, { "epoch": 0.4, "learning_rate": 3.298084752147487e-05, "loss": 0.016, "step": 920 }, { "epoch": 0.41, "learning_rate": 3.2653219522528106e-05, "loss": 0.03, "step": 930 }, { "epoch": 0.41, "learning_rate": 3.23241346865727e-05, "loss": 0.0025, "step": 940 }, { "epoch": 0.42, "learning_rate": 3.199365565691406e-05, "loss": 0.0752, "step": 950 }, { "epoch": 0.42, "learning_rate": 3.166184534225087e-05, "loss": 0.0078, "step": 960 }, { "epoch": 0.43, "learning_rate": 3.132876690470008e-05, "loss": 0.0326, "step": 970 }, { "epoch": 0.43, "learning_rate": 3.099448374777351e-05, "loss": 0.0249, "step": 980 }, { "epoch": 0.43, "learning_rate": 3.065905950430867e-05, "loss": 0.003, "step": 990 }, { "epoch": 0.44, "learning_rate": 3.032255802435588e-05, "loss": 0.0239, "step": 1000 }, { "epoch": 0.44, "learning_rate": 2.9985043363023958e-05, "loss": 0.0008, "step": 1010 }, { "epoch": 0.45, "learning_rate": 2.9646579768287002e-05, "loss": 0.0722, "step": 1020 }, { "epoch": 0.45, "learning_rate": 2.930723166875437e-05, "loss": 0.0007, "step": 1030 }, { "epoch": 0.46, "learning_rate": 2.8967063661406285e-05, "loss": 0.0246, "step": 1040 }, { "epoch": 0.46, "learning_rate": 2.862614049929743e-05, "loss": 0.0235, "step": 1050 }, { "epoch": 0.47, "learning_rate": 2.8284527079230822e-05, "loss": 0.0079, "step": 1060 }, { "epoch": 0.47, "learning_rate": 2.7942288429404256e-05, "loss": 0.0271, "step": 1070 }, { "epoch": 0.47, "learning_rate": 2.759948969703185e-05, "loss": 0.001, "step": 1080 }, { "epoch": 0.48, "learning_rate": 2.7256196135942836e-05, "loss": 0.0087, "step": 1090 }, { "epoch": 0.48, "learning_rate": 2.6912473094160106e-05, "loss": 0.0737, "step": 1100 }, { "epoch": 0.49, "learning_rate": 2.6568386001460854e-05, "loss": 0.0095, "step": 1110 }, { "epoch": 0.49, "learning_rate": 2.6224000356921546e-05, "loss": 0.0533, "step": 1120 }, { "epoch": 0.5, "learning_rate": 2.587938171644983e-05, "loss": 0.027, "step": 1130 }, { "epoch": 0.5, "learning_rate": 2.5534595680305544e-05, "loss": 0.024, "step": 1140 }, { "epoch": 0.5, "learning_rate": 2.5189707880613334e-05, "loss": 0.0246, "step": 1150 }, { "epoch": 0.51, "learning_rate": 2.4844783968869126e-05, "loss": 0.0445, "step": 1160 }, { "epoch": 0.51, "learning_rate": 2.4499889603443018e-05, "loss": 0.0377, "step": 1170 }, { "epoch": 0.52, "learning_rate": 2.4155090437080776e-05, "loss": 0.0124, "step": 1180 }, { "epoch": 0.52, "learning_rate": 2.3810452104406444e-05, "loss": 0.0351, "step": 1190 }, { "epoch": 0.53, "learning_rate": 2.3466040209428406e-05, "loss": 0.0352, "step": 1200 }, { "epoch": 0.53, "learning_rate": 2.312192031305123e-05, "loss": 0.0319, "step": 1210 }, { "epoch": 0.54, "learning_rate": 2.277815792059578e-05, "loss": 0.0201, "step": 1220 }, { "epoch": 0.54, "learning_rate": 2.243481846932987e-05, "loss": 0.0253, "step": 1230 }, { "epoch": 0.54, "learning_rate": 2.209196731601187e-05, "loss": 0.0368, "step": 1240 }, { "epoch": 0.55, "learning_rate": 2.174966972444967e-05, "loss": 0.002, "step": 1250 }, { "epoch": 0.55, "learning_rate": 2.1407990853077298e-05, "loss": 0.0541, "step": 1260 }, { "epoch": 0.56, "learning_rate": 2.106699574255162e-05, "loss": 0.0015, "step": 1270 }, { "epoch": 0.56, "learning_rate": 2.0726749303371447e-05, "loss": 0.0082, "step": 1280 }, { "epoch": 0.57, "learning_rate": 2.0387316303521454e-05, "loss": 0.039, "step": 1290 }, { "epoch": 0.57, "learning_rate": 2.0048761356143165e-05, "loss": 0.0139, "step": 1300 }, { "epoch": 0.58, "learning_rate": 1.9711148907235455e-05, "loss": 0.05, "step": 1310 }, { "epoch": 0.58, "learning_rate": 1.9374543223386888e-05, "loss": 0.0685, "step": 1320 }, { "epoch": 0.58, "learning_rate": 1.903900837954213e-05, "loss": 0.0047, "step": 1330 }, { "epoch": 0.59, "learning_rate": 1.8704608246804956e-05, "loss": 0.0069, "step": 1340 }, { "epoch": 0.59, "learning_rate": 1.8371406480279934e-05, "loss": 0.0138, "step": 1350 }, { "epoch": 0.6, "learning_rate": 1.8039466506955334e-05, "loss": 0.0057, "step": 1360 }, { "epoch": 0.6, "learning_rate": 1.7708851513629377e-05, "loss": 0.0021, "step": 1370 }, { "epoch": 0.61, "learning_rate": 1.737962443488228e-05, "loss": 0.0036, "step": 1380 }, { "epoch": 0.61, "learning_rate": 1.705184794109625e-05, "loss": 0.0485, "step": 1390 }, { "epoch": 0.61, "learning_rate": 1.6725584426525786e-05, "loss": 0.0006, "step": 1400 }, { "epoch": 0.62, "learning_rate": 1.6400895997420585e-05, "loss": 0.0179, "step": 1410 }, { "epoch": 0.62, "learning_rate": 1.6077844460203206e-05, "loss": 0.0009, "step": 1420 }, { "epoch": 0.63, "learning_rate": 1.5756491309703875e-05, "loss": 0.012, "step": 1430 }, { "epoch": 0.63, "learning_rate": 1.5436897717454524e-05, "loss": 0.0197, "step": 1440 }, { "epoch": 0.64, "learning_rate": 1.5119124520044461e-05, "loss": 0.0234, "step": 1450 }, { "epoch": 0.64, "learning_rate": 1.480323220753972e-05, "loss": 0.0023, "step": 1460 }, { "epoch": 0.65, "learning_rate": 1.4489280911968438e-05, "loss": 0.0252, "step": 1470 }, { "epoch": 0.65, "learning_rate": 1.4177330395874327e-05, "loss": 0.0014, "step": 1480 }, { "epoch": 0.65, "learning_rate": 1.386744004094052e-05, "loss": 0.0024, "step": 1490 }, { "epoch": 0.66, "learning_rate": 1.3559668836685946e-05, "loss": 0.0427, "step": 1500 }, { "epoch": 0.66, "learning_rate": 1.3254075369236238e-05, "loss": 0.0065, "step": 1510 }, { "epoch": 0.67, "learning_rate": 1.2950717810171558e-05, "loss": 0.0061, "step": 1520 }, { "epoch": 0.67, "learning_rate": 1.2649653905453284e-05, "loss": 0.0064, "step": 1530 }, { "epoch": 0.68, "learning_rate": 1.2350940964431687e-05, "loss": 0.0604, "step": 1540 }, { "epoch": 0.68, "learning_rate": 1.2054635848936771e-05, "loss": 0.0699, "step": 1550 }, { "epoch": 0.68, "learning_rate": 1.1760794962454257e-05, "loss": 0.0061, "step": 1560 }, { "epoch": 0.69, "learning_rate": 1.1469474239388858e-05, "loss": 0.0775, "step": 1570 }, { "epoch": 0.69, "learning_rate": 1.1180729134416796e-05, "loss": 0.0388, "step": 1580 }, { "epoch": 0.7, "learning_rate": 1.0894614611929702e-05, "loss": 0.0081, "step": 1590 }, { "epoch": 0.7, "learning_rate": 1.0611185135571777e-05, "loss": 0.0039, "step": 1600 }, { "epoch": 0.71, "learning_rate": 1.0330494657872312e-05, "loss": 0.0019, "step": 1610 }, { "epoch": 0.71, "learning_rate": 1.0052596609975507e-05, "loss": 0.1103, "step": 1620 }, { "epoch": 0.72, "learning_rate": 9.777543891469517e-06, "loss": 0.0232, "step": 1630 }, { "epoch": 0.72, "learning_rate": 9.5053888603167e-06, "loss": 0.0157, "step": 1640 }, { "epoch": 0.72, "learning_rate": 9.236183322886945e-06, "loss": 0.0028, "step": 1650 }, { "epoch": 0.73, "learning_rate": 8.969978524096034e-06, "loss": 0.0009, "step": 1660 }, { "epoch": 0.73, "learning_rate": 8.706825137650837e-06, "loss": 0.0138, "step": 1670 }, { "epoch": 0.74, "learning_rate": 8.446773256403298e-06, "loss": 0.0409, "step": 1680 }, { "epoch": 0.74, "learning_rate": 8.189872382814912e-06, "loss": 0.0603, "step": 1690 }, { "epoch": 0.75, "learning_rate": 7.936171419533653e-06, "loss": 0.0158, "step": 1700 }, { "epoch": 0.75, "learning_rate": 7.685718660085057e-06, "loss": 0.0167, "step": 1710 }, { "epoch": 0.76, "learning_rate": 7.4385617796792516e-06, "loss": 0.0161, "step": 1720 }, { "epoch": 0.76, "learning_rate": 7.194747826135687e-06, "loss": 0.0015, "step": 1730 }, { "epoch": 0.76, "learning_rate": 6.954323210927299e-06, "loss": 0.0633, "step": 1740 }, { "epoch": 0.77, "learning_rate": 6.717333700345807e-06, "loss": 0.0544, "step": 1750 }, { "epoch": 0.77, "learning_rate": 6.483824406789798e-06, "loss": 0.0022, "step": 1760 }, { "epoch": 0.78, "learning_rate": 6.2538397801773594e-06, "loss": 0.0341, "step": 1770 }, { "epoch": 0.78, "learning_rate": 6.027423599484697e-06, "loss": 0.0167, "step": 1780 }, { "epoch": 0.79, "learning_rate": 5.804618964412586e-06, "loss": 0.024, "step": 1790 }, { "epoch": 0.79, "learning_rate": 5.585468287182071e-06, "loss": 0.0293, "step": 1800 }, { "epoch": 0.79, "learning_rate": 5.391391304955415e-06, "loss": 0.0008, "step": 1810 }, { "epoch": 0.8, "learning_rate": 5.1792974957679375e-06, "loss": 0.0123, "step": 1820 }, { "epoch": 0.8, "learning_rate": 4.970976678177083e-06, "loss": 0.0024, "step": 1830 }, { "epoch": 0.81, "learning_rate": 4.766468507320909e-06, "loss": 0.024, "step": 1840 }, { "epoch": 0.81, "learning_rate": 4.565811912576912e-06, "loss": 0.0593, "step": 1850 }, { "epoch": 0.82, "learning_rate": 4.369045090151591e-06, "loss": 0.0164, "step": 1860 }, { "epoch": 0.82, "learning_rate": 4.176205495809563e-06, "loss": 0.0213, "step": 1870 }, { "epoch": 0.83, "learning_rate": 3.987329837743625e-06, "loss": 0.0155, "step": 1880 }, { "epoch": 0.83, "learning_rate": 3.8024540695871274e-06, "loss": 0.0038, "step": 1890 }, { "epoch": 0.83, "learning_rate": 3.6216133835699995e-06, "loss": 0.0519, "step": 1900 }, { "epoch": 0.84, "learning_rate": 3.444842203819662e-06, "loss": 0.0105, "step": 1910 }, { "epoch": 0.84, "learning_rate": 3.2721741798082073e-06, "loss": 0.0279, "step": 1920 }, { "epoch": 0.85, "learning_rate": 3.1036421799470057e-06, "loss": 0.0315, "step": 1930 }, { "epoch": 0.85, "learning_rate": 2.9392782853300077e-06, "loss": 0.0124, "step": 1940 }, { "epoch": 0.86, "learning_rate": 2.779113783626916e-06, "loss": 0.0008, "step": 1950 }, { "epoch": 0.86, "learning_rate": 2.623179163127379e-06, "loss": 0.0012, "step": 1960 }, { "epoch": 0.86, "learning_rate": 2.471504106937364e-06, "loss": 0.0241, "step": 1970 }, { "epoch": 0.87, "learning_rate": 2.324117487328789e-06, "loss": 0.027, "step": 1980 }, { "epoch": 0.87, "learning_rate": 2.1810473602435324e-06, "loss": 0.0752, "step": 1990 }, { "epoch": 0.88, "learning_rate": 2.0423209599527814e-06, "loss": 0.0041, "step": 2000 } ], "max_steps": 2277, "num_train_epochs": 1, "total_flos": 0.0, "trial_name": null, "trial_params": null }