{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9832036050798852, "global_step": 3000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 4.999915179102931e-05, "loss": 2.5998, "step": 10 }, { "epoch": 0.01, "learning_rate": 4.999570604073014e-05, "loss": 2.4323, "step": 20 }, { "epoch": 0.01, "learning_rate": 4.998961010109915e-05, "loss": 2.2223, "step": 30 }, { "epoch": 0.01, "learning_rate": 4.9980864618463135e-05, "loss": 2.2165, "step": 40 }, { "epoch": 0.02, "learning_rate": 4.996947052006874e-05, "loss": 2.1109, "step": 50 }, { "epoch": 0.02, "learning_rate": 4.9955429013984174e-05, "loss": 2.0959, "step": 60 }, { "epoch": 0.02, "learning_rate": 4.993874158897114e-05, "loss": 2.094, "step": 70 }, { "epoch": 0.03, "learning_rate": 4.9919410014326955e-05, "loss": 2.0554, "step": 80 }, { "epoch": 0.03, "learning_rate": 4.989743633969698e-05, "loss": 2.0815, "step": 90 }, { "epoch": 0.03, "learning_rate": 4.9872822894857296e-05, "loss": 2.1007, "step": 100 }, { "epoch": 0.04, "learning_rate": 4.98455722894677e-05, "loss": 2.1199, "step": 110 }, { "epoch": 0.04, "learning_rate": 4.981568741279498e-05, "loss": 2.1445, "step": 120 }, { "epoch": 0.04, "learning_rate": 4.978317143340662e-05, "loss": 2.0086, "step": 130 }, { "epoch": 0.05, "learning_rate": 4.974802779883484e-05, "loss": 2.0287, "step": 140 }, { "epoch": 0.05, "learning_rate": 4.971026023521105e-05, "loss": 2.0494, "step": 150 }, { "epoch": 0.05, "learning_rate": 4.966987274687078e-05, "loss": 1.971, "step": 160 }, { "epoch": 0.06, "learning_rate": 4.9626869615929175e-05, "loss": 2.1144, "step": 170 }, { "epoch": 0.06, "learning_rate": 4.958125540182689e-05, "loss": 2.0514, "step": 180 }, { "epoch": 0.06, "learning_rate": 4.953303494084675e-05, "loss": 2.0357, "step": 190 }, { "epoch": 0.07, "learning_rate": 4.9482213345600936e-05, "loss": 2.0314, "step": 200 }, { "epoch": 0.07, "learning_rate": 4.942879600448893e-05, "loss": 1.9785, "step": 210 }, { "epoch": 0.07, "learning_rate": 4.9372788581126215e-05, "loss": 2.1087, "step": 220 }, { "epoch": 0.08, "learning_rate": 4.931419701374377e-05, "loss": 2.0925, "step": 230 }, { "epoch": 0.08, "learning_rate": 4.925302751455846e-05, "loss": 2.1145, "step": 240 }, { "epoch": 0.08, "learning_rate": 4.918928656911439e-05, "loss": 1.9985, "step": 250 }, { "epoch": 0.09, "learning_rate": 4.91229809355953e-05, "loss": 1.939, "step": 260 }, { "epoch": 0.09, "learning_rate": 4.905411764410796e-05, "loss": 2.0781, "step": 270 }, { "epoch": 0.09, "learning_rate": 4.898270399593684e-05, "loss": 2.1217, "step": 280 }, { "epoch": 0.1, "learning_rate": 4.8908747562769995e-05, "loss": 2.0891, "step": 290 }, { "epoch": 0.1, "learning_rate": 4.8832256185896244e-05, "loss": 2.0106, "step": 300 }, { "epoch": 0.1, "learning_rate": 4.875323797537378e-05, "loss": 2.0148, "step": 310 }, { "epoch": 0.1, "learning_rate": 4.867170130917034e-05, "loss": 2.0716, "step": 320 }, { "epoch": 0.11, "learning_rate": 4.858765483227488e-05, "loss": 2.1095, "step": 330 }, { "epoch": 0.11, "learning_rate": 4.8501107455781024e-05, "loss": 2.0128, "step": 340 }, { "epoch": 0.11, "learning_rate": 4.841206835594222e-05, "loss": 2.0141, "step": 350 }, { "epoch": 0.12, "learning_rate": 4.832054697319884e-05, "loss": 2.0029, "step": 360 }, { "epoch": 0.12, "learning_rate": 4.8226553011177265e-05, "loss": 2.013, "step": 370 }, { "epoch": 0.12, "learning_rate": 4.813009643566101e-05, "loss": 1.9818, "step": 380 }, { "epoch": 0.13, "learning_rate": 4.803118747353412e-05, "loss": 1.9709, "step": 390 }, { "epoch": 0.13, "learning_rate": 4.792983661169685e-05, "loss": 2.0799, "step": 400 }, { "epoch": 0.13, "learning_rate": 4.7826054595953815e-05, "loss": 2.042, "step": 410 }, { "epoch": 0.14, "learning_rate": 4.77198524298746e-05, "loss": 1.9697, "step": 420 }, { "epoch": 0.14, "learning_rate": 4.7611241373627136e-05, "loss": 2.0052, "step": 430 }, { "epoch": 0.14, "learning_rate": 4.750023294278381e-05, "loss": 2.0353, "step": 440 }, { "epoch": 0.15, "learning_rate": 4.738683890710056e-05, "loss": 2.0077, "step": 450 }, { "epoch": 0.15, "learning_rate": 4.727107128926892e-05, "loss": 1.984, "step": 460 }, { "epoch": 0.15, "learning_rate": 4.715294236364135e-05, "loss": 1.9694, "step": 470 }, { "epoch": 0.16, "learning_rate": 4.70324646549298e-05, "loss": 1.9918, "step": 480 }, { "epoch": 0.16, "learning_rate": 4.690965093687781e-05, "loss": 2.0472, "step": 490 }, { "epoch": 0.16, "learning_rate": 4.6784514230906146e-05, "loss": 1.8565, "step": 500 }, { "epoch": 0.17, "learning_rate": 4.6657067804732156e-05, "loss": 2.0349, "step": 510 }, { "epoch": 0.17, "learning_rate": 4.652732517096311e-05, "loss": 2.0547, "step": 520 }, { "epoch": 0.17, "learning_rate": 4.639530008566349e-05, "loss": 2.0175, "step": 530 }, { "epoch": 0.18, "learning_rate": 4.626100654689648e-05, "loss": 2.0696, "step": 540 }, { "epoch": 0.18, "learning_rate": 4.612445879323982e-05, "loss": 2.0239, "step": 550 }, { "epoch": 0.18, "learning_rate": 4.5985671302276165e-05, "loss": 2.1, "step": 560 }, { "epoch": 0.19, "learning_rate": 4.584465878905807e-05, "loss": 1.9731, "step": 570 }, { "epoch": 0.19, "learning_rate": 4.570143620454781e-05, "loss": 1.9604, "step": 580 }, { "epoch": 0.19, "learning_rate": 4.55560187340322e-05, "loss": 1.9924, "step": 590 }, { "epoch": 0.2, "learning_rate": 4.5408421795512576e-05, "loss": 2.049, "step": 600 }, { "epoch": 0.2, "learning_rate": 4.525866103807007e-05, "loss": 1.9764, "step": 610 }, { "epoch": 0.2, "learning_rate": 4.5106752340206435e-05, "loss": 1.9687, "step": 620 }, { "epoch": 0.21, "learning_rate": 4.495271180816045e-05, "loss": 2.0358, "step": 630 }, { "epoch": 0.21, "learning_rate": 4.479655577420031e-05, "loss": 2.076, "step": 640 }, { "epoch": 0.21, "learning_rate": 4.463830079489196e-05, "loss": 2.0922, "step": 650 }, { "epoch": 0.22, "learning_rate": 4.447796364934368e-05, "loss": 1.951, "step": 660 }, { "epoch": 0.22, "learning_rate": 4.431556133742702e-05, "loss": 1.936, "step": 670 }, { "epoch": 0.22, "learning_rate": 4.415111107797445e-05, "loss": 1.9275, "step": 680 }, { "epoch": 0.23, "learning_rate": 4.3984630306953676e-05, "loss": 2.0845, "step": 690 }, { "epoch": 0.23, "learning_rate": 4.381613667561898e-05, "loss": 2.0253, "step": 700 }, { "epoch": 0.23, "learning_rate": 4.3645648048639734e-05, "loss": 1.9797, "step": 710 }, { "epoch": 0.24, "learning_rate": 4.34731825022063e-05, "loss": 1.9952, "step": 720 }, { "epoch": 0.24, "learning_rate": 4.3298758322113455e-05, "loss": 1.9893, "step": 730 }, { "epoch": 0.24, "learning_rate": 4.312239400182166e-05, "loss": 2.006, "step": 740 }, { "epoch": 0.25, "learning_rate": 4.294410824049626e-05, "loss": 1.9752, "step": 750 }, { "epoch": 0.25, "learning_rate": 4.2763919941024904e-05, "loss": 2.0759, "step": 760 }, { "epoch": 0.25, "learning_rate": 4.258184820801331e-05, "loss": 1.973, "step": 770 }, { "epoch": 0.26, "learning_rate": 4.239791234575974e-05, "loss": 1.9161, "step": 780 }, { "epoch": 0.26, "learning_rate": 4.2212131856208205e-05, "loss": 2.0307, "step": 790 }, { "epoch": 0.26, "learning_rate": 4.2024526436880785e-05, "loss": 1.9427, "step": 800 }, { "epoch": 0.27, "learning_rate": 4.1835115978789156e-05, "loss": 1.9997, "step": 810 }, { "epoch": 0.27, "learning_rate": 4.164392056432566e-05, "loss": 2.069, "step": 820 }, { "epoch": 0.27, "learning_rate": 4.1450960465134025e-05, "loss": 1.9659, "step": 830 }, { "epoch": 0.28, "learning_rate": 4.1256256139960065e-05, "loss": 1.9834, "step": 840 }, { "epoch": 0.28, "learning_rate": 4.1059828232482524e-05, "loss": 1.9762, "step": 850 }, { "epoch": 0.28, "learning_rate": 4.086169756912431e-05, "loss": 1.9506, "step": 860 }, { "epoch": 0.29, "learning_rate": 4.0661885156844356e-05, "loss": 2.0073, "step": 870 }, { "epoch": 0.29, "learning_rate": 4.0460412180910336e-05, "loss": 1.9905, "step": 880 }, { "epoch": 0.29, "learning_rate": 4.025730000265251e-05, "loss": 2.0375, "step": 890 }, { "epoch": 0.29, "learning_rate": 4.0052570157198833e-05, "loss": 2.0074, "step": 900 }, { "epoch": 0.3, "learning_rate": 3.984624435119168e-05, "loss": 2.089, "step": 910 }, { "epoch": 0.3, "learning_rate": 3.963834446048644e-05, "loss": 2.044, "step": 920 }, { "epoch": 0.3, "learning_rate": 3.942889252783202e-05, "loss": 1.9993, "step": 930 }, { "epoch": 0.31, "learning_rate": 3.921791076053379e-05, "loss": 1.9848, "step": 940 }, { "epoch": 0.31, "learning_rate": 3.90054215280991e-05, "loss": 1.954, "step": 950 }, { "epoch": 0.31, "learning_rate": 3.879144735986539e-05, "loss": 2.0205, "step": 960 }, { "epoch": 0.32, "learning_rate": 3.8576010942611635e-05, "loss": 2.0108, "step": 970 }, { "epoch": 0.32, "learning_rate": 3.83591351181529e-05, "loss": 1.9914, "step": 980 }, { "epoch": 0.32, "learning_rate": 3.8140842880918515e-05, "loss": 1.9914, "step": 990 }, { "epoch": 0.33, "learning_rate": 3.792115737551411e-05, "loss": 1.8875, "step": 1000 }, { "epoch": 0.33, "learning_rate": 3.7700101894267615e-05, "loss": 2.0138, "step": 1010 }, { "epoch": 0.33, "learning_rate": 3.7477699874759784e-05, "loss": 1.9455, "step": 1020 }, { "epoch": 0.34, "learning_rate": 3.725397489733912e-05, "loss": 1.9611, "step": 1030 }, { "epoch": 0.34, "learning_rate": 3.702895068262177e-05, "loss": 1.9679, "step": 1040 }, { "epoch": 0.34, "learning_rate": 3.680265108897652e-05, "loss": 2.0271, "step": 1050 }, { "epoch": 0.35, "learning_rate": 3.657510010999524e-05, "loss": 1.9258, "step": 1060 }, { "epoch": 0.35, "learning_rate": 3.63463218719489e-05, "loss": 2.0162, "step": 1070 }, { "epoch": 0.35, "learning_rate": 3.6116340631229574e-05, "loss": 1.9973, "step": 1080 }, { "epoch": 0.36, "learning_rate": 3.588518077177861e-05, "loss": 2.0753, "step": 1090 }, { "epoch": 0.36, "learning_rate": 3.565286680250138e-05, "loss": 2.0728, "step": 1100 }, { "epoch": 0.36, "learning_rate": 3.541942335466864e-05, "loss": 1.993, "step": 1110 }, { "epoch": 0.37, "learning_rate": 3.518487517930495e-05, "loss": 1.9577, "step": 1120 }, { "epoch": 0.37, "learning_rate": 3.494924714456454e-05, "loss": 1.9794, "step": 1130 }, { "epoch": 0.37, "learning_rate": 3.471256423309454e-05, "loss": 2.0123, "step": 1140 }, { "epoch": 0.38, "learning_rate": 3.4474851539386214e-05, "loss": 2.0714, "step": 1150 }, { "epoch": 0.38, "learning_rate": 3.423613426711431e-05, "loss": 2.0283, "step": 1160 }, { "epoch": 0.38, "learning_rate": 3.3996437726464787e-05, "loss": 2.0507, "step": 1170 }, { "epoch": 0.39, "learning_rate": 3.375578733145131e-05, "loss": 1.9589, "step": 1180 }, { "epoch": 0.39, "learning_rate": 3.3514208597220705e-05, "loss": 2.0361, "step": 1190 }, { "epoch": 0.39, "learning_rate": 3.327172713734768e-05, "loss": 2.0543, "step": 1200 }, { "epoch": 0.4, "learning_rate": 3.3028368661119166e-05, "loss": 1.963, "step": 1210 }, { "epoch": 0.4, "learning_rate": 3.278415897080839e-05, "loss": 2.0818, "step": 1220 }, { "epoch": 0.4, "learning_rate": 3.253912395893927e-05, "loss": 1.9921, "step": 1230 }, { "epoch": 0.41, "learning_rate": 3.2293289605541063e-05, "loss": 2.0351, "step": 1240 }, { "epoch": 0.41, "learning_rate": 3.204668197539385e-05, "loss": 2.0442, "step": 1250 }, { "epoch": 0.41, "learning_rate": 3.179932721526497e-05, "loss": 2.024, "step": 1260 }, { "epoch": 0.42, "learning_rate": 3.155125155113682e-05, "loss": 2.0295, "step": 1270 }, { "epoch": 0.42, "learning_rate": 3.13024812854262e-05, "loss": 2.0086, "step": 1280 }, { "epoch": 0.42, "learning_rate": 3.105304279419559e-05, "loss": 2.0039, "step": 1290 }, { "epoch": 0.43, "learning_rate": 3.0802962524356606e-05, "loss": 2.0377, "step": 1300 }, { "epoch": 0.43, "learning_rate": 3.0552266990865934e-05, "loss": 1.946, "step": 1310 }, { "epoch": 0.43, "learning_rate": 3.0300982773914093e-05, "loss": 2.0119, "step": 1320 }, { "epoch": 0.44, "learning_rate": 3.0049136516107205e-05, "loss": 1.9564, "step": 1330 }, { "epoch": 0.44, "learning_rate": 2.9796754919642228e-05, "loss": 1.9831, "step": 1340 }, { "epoch": 0.44, "learning_rate": 2.954386474347582e-05, "loss": 1.945, "step": 1350 }, { "epoch": 0.45, "learning_rate": 2.929049280048719e-05, "loss": 1.9298, "step": 1360 }, { "epoch": 0.45, "learning_rate": 2.9036665954635266e-05, "loss": 1.939, "step": 1370 }, { "epoch": 0.45, "learning_rate": 2.8782411118110397e-05, "loss": 2.0291, "step": 1380 }, { "epoch": 0.46, "learning_rate": 2.852775524848098e-05, "loss": 1.9421, "step": 1390 }, { "epoch": 0.46, "learning_rate": 2.8272725345835278e-05, "loss": 1.9836, "step": 1400 }, { "epoch": 0.46, "learning_rate": 2.8017348449918672e-05, "loss": 2.0366, "step": 1410 }, { "epoch": 0.47, "learning_rate": 2.7761651637266795e-05, "loss": 2.0529, "step": 1420 }, { "epoch": 0.47, "learning_rate": 2.7505662018334698e-05, "loss": 2.0653, "step": 1430 }, { "epoch": 0.47, "learning_rate": 2.7249406734622446e-05, "loss": 1.938, "step": 1440 }, { "epoch": 0.48, "learning_rate": 2.699291295579743e-05, "loss": 1.9879, "step": 1450 }, { "epoch": 0.48, "learning_rate": 2.6736207876813646e-05, "loss": 2.0756, "step": 1460 }, { "epoch": 0.48, "learning_rate": 2.647931871502839e-05, "loss": 2.0165, "step": 1470 }, { "epoch": 0.49, "learning_rate": 2.6222272707316453e-05, "loss": 1.9832, "step": 1480 }, { "epoch": 0.49, "learning_rate": 2.596509710718235e-05, "loss": 2.0621, "step": 1490 }, { "epoch": 0.49, "learning_rate": 2.570781918187074e-05, "loss": 2.0004, "step": 1500 }, { "epoch": 0.49, "learning_rate": 2.545046620947538e-05, "loss": 1.9492, "step": 1510 }, { "epoch": 0.5, "learning_rate": 2.5193065476046955e-05, "loss": 2.0872, "step": 1520 }, { "epoch": 0.5, "learning_rate": 2.4935644272700044e-05, "loss": 1.9346, "step": 1530 }, { "epoch": 0.5, "learning_rate": 2.4678229892719587e-05, "loss": 1.9643, "step": 1540 }, { "epoch": 0.51, "learning_rate": 2.4420849628667046e-05, "loss": 1.9377, "step": 1550 }, { "epoch": 0.51, "learning_rate": 2.4163530769486737e-05, "loss": 2.002, "step": 1560 }, { "epoch": 0.51, "learning_rate": 2.3906300597612456e-05, "loss": 1.9659, "step": 1570 }, { "epoch": 0.52, "learning_rate": 2.3649186386074872e-05, "loss": 1.9492, "step": 1580 }, { "epoch": 0.52, "learning_rate": 2.3392215395609862e-05, "loss": 1.9525, "step": 1590 }, { "epoch": 0.52, "learning_rate": 2.313541487176818e-05, "loss": 2.013, "step": 1600 }, { "epoch": 0.53, "learning_rate": 2.2878812042026713e-05, "loss": 2.0389, "step": 1610 }, { "epoch": 0.53, "learning_rate": 2.2622434112901694e-05, "loss": 1.9962, "step": 1620 }, { "epoch": 0.53, "learning_rate": 2.2366308267064072e-05, "loss": 1.9724, "step": 1630 }, { "epoch": 0.54, "learning_rate": 2.2110461660457482e-05, "loss": 2.0942, "step": 1640 }, { "epoch": 0.54, "learning_rate": 2.1854921419418996e-05, "loss": 1.9673, "step": 1650 }, { "epoch": 0.54, "learning_rate": 2.159971463780304e-05, "loss": 1.969, "step": 1660 }, { "epoch": 0.55, "learning_rate": 2.1344868374108732e-05, "loss": 2.0669, "step": 1670 }, { "epoch": 0.55, "learning_rate": 2.1090409648611005e-05, "loss": 2.069, "step": 1680 }, { "epoch": 0.55, "learning_rate": 2.0836365440495754e-05, "loss": 1.9849, "step": 1690 }, { "epoch": 0.56, "learning_rate": 2.0582762684999327e-05, "loss": 2.0086, "step": 1700 }, { "epoch": 0.56, "learning_rate": 2.032962827055273e-05, "loss": 1.9318, "step": 1710 }, { "epoch": 0.56, "learning_rate": 2.0076989035930732e-05, "loss": 1.9637, "step": 1720 }, { "epoch": 0.57, "learning_rate": 1.982487176740627e-05, "loss": 2.0757, "step": 1730 }, { "epoch": 0.57, "learning_rate": 1.957330319591043e-05, "loss": 2.0004, "step": 1740 }, { "epoch": 0.57, "learning_rate": 1.9322309994198255e-05, "loss": 1.9287, "step": 1750 }, { "epoch": 0.58, "learning_rate": 1.907191877402076e-05, "loss": 1.9431, "step": 1760 }, { "epoch": 0.58, "learning_rate": 1.8822156083303387e-05, "loss": 1.8691, "step": 1770 }, { "epoch": 0.58, "learning_rate": 1.8573048403331256e-05, "loss": 1.9908, "step": 1780 }, { "epoch": 0.59, "learning_rate": 1.832462214594146e-05, "loss": 2.0072, "step": 1790 }, { "epoch": 0.59, "learning_rate": 1.8076903650722733e-05, "loss": 1.9912, "step": 1800 }, { "epoch": 0.59, "learning_rate": 1.7829919182222752e-05, "loss": 1.9576, "step": 1810 }, { "epoch": 0.6, "learning_rate": 1.7583694927163453e-05, "loss": 2.0165, "step": 1820 }, { "epoch": 0.6, "learning_rate": 1.7338256991664537e-05, "loss": 1.9341, "step": 1830 }, { "epoch": 0.6, "learning_rate": 1.709363139847557e-05, "loss": 1.9782, "step": 1840 }, { "epoch": 0.61, "learning_rate": 1.6849844084216874e-05, "loss": 1.9201, "step": 1850 }, { "epoch": 0.61, "learning_rate": 1.660692089662959e-05, "loss": 1.932, "step": 1860 }, { "epoch": 0.61, "learning_rate": 1.6364887591835155e-05, "loss": 1.9732, "step": 1870 }, { "epoch": 0.62, "learning_rate": 1.612376983160449e-05, "loss": 1.9243, "step": 1880 }, { "epoch": 0.62, "learning_rate": 1.5883593180637168e-05, "loss": 2.0249, "step": 1890 }, { "epoch": 0.62, "learning_rate": 1.5644383103850944e-05, "loss": 1.9553, "step": 1900 }, { "epoch": 0.63, "learning_rate": 1.5406164963681774e-05, "loss": 2.0105, "step": 1910 }, { "epoch": 0.63, "learning_rate": 1.5168964017394738e-05, "loss": 2.0154, "step": 1920 }, { "epoch": 0.63, "learning_rate": 1.4932805414406164e-05, "loss": 2.0062, "step": 1930 }, { "epoch": 0.64, "learning_rate": 1.4697714193617097e-05, "loss": 1.998, "step": 1940 }, { "epoch": 0.64, "learning_rate": 1.446371528075855e-05, "loss": 1.9942, "step": 1950 }, { "epoch": 0.64, "learning_rate": 1.423083348574874e-05, "loss": 1.9284, "step": 1960 }, { "epoch": 0.65, "learning_rate": 1.3999093500062587e-05, "loss": 1.9713, "step": 1970 }, { "epoch": 0.65, "learning_rate": 1.3768519894113773e-05, "loss": 1.8622, "step": 1980 }, { "epoch": 0.65, "learning_rate": 1.3539137114649664e-05, "loss": 1.97, "step": 1990 }, { "epoch": 0.66, "learning_rate": 1.3310969482159297e-05, "loss": 1.9003, "step": 2000 }, { "epoch": 0.66, "learning_rate": 1.308404118829482e-05, "loss": 1.9686, "step": 2010 }, { "epoch": 0.66, "learning_rate": 1.2858376293306523e-05, "loss": 2.0165, "step": 2020 }, { "epoch": 0.67, "learning_rate": 1.263399872349183e-05, "loss": 2.0162, "step": 2030 }, { "epoch": 0.67, "learning_rate": 1.2410932268658521e-05, "loss": 1.9395, "step": 2040 }, { "epoch": 0.67, "learning_rate": 1.2189200579602369e-05, "loss": 1.9701, "step": 2050 }, { "epoch": 0.68, "learning_rate": 1.1968827165599564e-05, "loss": 1.9819, "step": 2060 }, { "epoch": 0.68, "learning_rate": 1.1749835391914121e-05, "loss": 2.0189, "step": 2070 }, { "epoch": 0.68, "learning_rate": 1.1532248477320548e-05, "loss": 2.0084, "step": 2080 }, { "epoch": 0.68, "learning_rate": 1.1316089491642076e-05, "loss": 1.975, "step": 2090 }, { "epoch": 0.69, "learning_rate": 1.110138135330464e-05, "loss": 1.9917, "step": 2100 }, { "epoch": 0.69, "learning_rate": 1.0888146826906976e-05, "loss": 1.9626, "step": 2110 }, { "epoch": 0.69, "learning_rate": 1.067640852080693e-05, "loss": 2.0499, "step": 2120 }, { "epoch": 0.7, "learning_rate": 1.0466188884724428e-05, "loss": 2.0384, "step": 2130 }, { "epoch": 0.7, "learning_rate": 1.0257510207361215e-05, "loss": 2.0733, "step": 2140 }, { "epoch": 0.7, "learning_rate": 1.0050394614037687e-05, "loss": 2.0521, "step": 2150 }, { "epoch": 0.71, "learning_rate": 9.844864064347015e-06, "loss": 1.8643, "step": 2160 }, { "epoch": 0.71, "learning_rate": 9.640940349826885e-06, "loss": 2.0635, "step": 2170 }, { "epoch": 0.71, "learning_rate": 9.438645091649029e-06, "loss": 1.987, "step": 2180 }, { "epoch": 0.72, "learning_rate": 9.237999738326812e-06, "loss": 1.9523, "step": 2190 }, { "epoch": 0.72, "learning_rate": 9.039025563441137e-06, "loss": 1.8975, "step": 2200 }, { "epoch": 0.72, "learning_rate": 8.841743663384913e-06, "loss": 2.0279, "step": 2210 }, { "epoch": 0.73, "learning_rate": 8.646174955126285e-06, "loss": 2.0118, "step": 2220 }, { "epoch": 0.73, "learning_rate": 8.452340173990882e-06, "loss": 1.9109, "step": 2230 }, { "epoch": 0.73, "learning_rate": 8.260259871463352e-06, "loss": 2.0255, "step": 2240 }, { "epoch": 0.74, "learning_rate": 8.0699544130084e-06, "loss": 2.0055, "step": 2250 }, { "epoch": 0.74, "learning_rate": 7.881443975911456e-06, "loss": 1.9446, "step": 2260 }, { "epoch": 0.74, "learning_rate": 7.69474854713943e-06, "loss": 2.0647, "step": 2270 }, { "epoch": 0.75, "learning_rate": 7.509887921221542e-06, "loss": 1.9681, "step": 2280 }, { "epoch": 0.75, "learning_rate": 7.3268816981506e-06, "loss": 1.9948, "step": 2290 }, { "epoch": 0.75, "learning_rate": 7.1457492813049e-06, "loss": 2.0114, "step": 2300 }, { "epoch": 0.76, "learning_rate": 6.966509875390961e-06, "loss": 1.8271, "step": 2310 }, { "epoch": 0.76, "learning_rate": 6.789182484407328e-06, "loss": 1.9599, "step": 2320 }, { "epoch": 0.76, "learning_rate": 6.613785909629686e-06, "loss": 1.9513, "step": 2330 }, { "epoch": 0.77, "learning_rate": 6.440338747617391e-06, "loss": 2.0155, "step": 2340 }, { "epoch": 0.77, "learning_rate": 6.268859388241799e-06, "loss": 1.9281, "step": 2350 }, { "epoch": 0.77, "learning_rate": 6.099366012736438e-06, "loss": 1.9911, "step": 2360 }, { "epoch": 0.78, "learning_rate": 5.931876591769345e-06, "loss": 1.8573, "step": 2370 }, { "epoch": 0.78, "learning_rate": 5.766408883537702e-06, "loss": 2.0132, "step": 2380 }, { "epoch": 0.78, "learning_rate": 5.602980431885024e-06, "loss": 2.0155, "step": 2390 }, { "epoch": 0.79, "learning_rate": 5.4416085644410304e-06, "loss": 2.0248, "step": 2400 }, { "epoch": 0.79, "learning_rate": 5.282310390784503e-06, "loss": 1.8942, "step": 2410 }, { "epoch": 0.79, "learning_rate": 5.125102800629203e-06, "loss": 2.0011, "step": 2420 }, { "epoch": 0.8, "learning_rate": 4.970002462033147e-06, "loss": 2.001, "step": 2430 }, { "epoch": 0.8, "learning_rate": 4.817025819631341e-06, "loss": 1.9498, "step": 2440 }, { "epoch": 0.8, "learning_rate": 4.666189092892245e-06, "loss": 1.9387, "step": 2450 }, { "epoch": 0.81, "learning_rate": 4.517508274398083e-06, "loss": 2.0008, "step": 2460 }, { "epoch": 0.81, "learning_rate": 4.370999128149211e-06, "loss": 1.912, "step": 2470 }, { "epoch": 0.81, "learning_rate": 4.226677187892739e-06, "loss": 2.0202, "step": 2480 }, { "epoch": 0.82, "learning_rate": 4.084557755475543e-06, "loss": 1.9186, "step": 2490 }, { "epoch": 0.82, "learning_rate": 3.94465589922188e-06, "loss": 1.9801, "step": 2500 }, { "epoch": 0.82, "learning_rate": 3.806986452335748e-06, "loss": 1.9139, "step": 2510 }, { "epoch": 0.83, "learning_rate": 3.67156401132821e-06, "loss": 2.0275, "step": 2520 }, { "epoch": 0.83, "learning_rate": 3.538402934469753e-06, "loss": 1.8819, "step": 2530 }, { "epoch": 0.83, "learning_rate": 3.4075173402679574e-06, "loss": 1.9807, "step": 2540 }, { "epoch": 0.84, "learning_rate": 3.2789211059705755e-06, "loss": 2.018, "step": 2550 }, { "epoch": 0.84, "learning_rate": 3.1526278660941792e-06, "loss": 1.9896, "step": 2560 }, { "epoch": 0.84, "learning_rate": 3.028651010978556e-06, "loss": 2.0217, "step": 2570 }, { "epoch": 0.85, "learning_rate": 2.9070036853669715e-06, "loss": 1.9522, "step": 2580 }, { "epoch": 0.85, "learning_rate": 2.7876987870125106e-06, "loss": 1.9938, "step": 2590 }, { "epoch": 0.85, "learning_rate": 2.67074896531056e-06, "loss": 1.9406, "step": 2600 }, { "epoch": 0.86, "learning_rate": 2.5561666199576757e-06, "loss": 1.9142, "step": 2610 }, { "epoch": 0.86, "learning_rate": 2.4439638996368585e-06, "loss": 1.9649, "step": 2620 }, { "epoch": 0.86, "learning_rate": 2.334152700729511e-06, "loss": 1.9985, "step": 2630 }, { "epoch": 0.87, "learning_rate": 2.226744666054098e-06, "loss": 2.006, "step": 2640 }, { "epoch": 0.87, "learning_rate": 2.1217511836317123e-06, "loss": 1.9711, "step": 2650 }, { "epoch": 0.87, "learning_rate": 2.019183385478654e-06, "loss": 2.0882, "step": 2660 }, { "epoch": 0.88, "learning_rate": 1.9190521464261425e-06, "loss": 1.9489, "step": 2670 }, { "epoch": 0.88, "learning_rate": 1.8213680829673092e-06, "loss": 1.993, "step": 2680 }, { "epoch": 0.88, "learning_rate": 1.7261415521315798e-06, "loss": 2.0167, "step": 2690 }, { "epoch": 0.88, "learning_rate": 1.6333826503865474e-06, "loss": 1.9981, "step": 2700 }, { "epoch": 0.89, "learning_rate": 1.543101212567502e-06, "loss": 1.9637, "step": 2710 }, { "epoch": 0.89, "learning_rate": 1.455306810834678e-06, "loss": 1.9172, "step": 2720 }, { "epoch": 0.89, "learning_rate": 1.3700087536583529e-06, "loss": 1.9828, "step": 2730 }, { "epoch": 0.9, "learning_rate": 1.287216084831927e-06, "loss": 1.9952, "step": 2740 }, { "epoch": 0.9, "learning_rate": 1.2069375825130292e-06, "loss": 1.9243, "step": 2750 }, { "epoch": 0.9, "learning_rate": 1.1291817582928104e-06, "loss": 1.9311, "step": 2760 }, { "epoch": 0.91, "learning_rate": 1.0539568562935003e-06, "loss": 1.9303, "step": 2770 }, { "epoch": 0.91, "learning_rate": 9.812708522943164e-07, "loss": 1.995, "step": 2780 }, { "epoch": 0.91, "learning_rate": 9.1113145288583e-07, "loss": 1.98, "step": 2790 }, { "epoch": 0.92, "learning_rate": 8.435460946528501e-07, "loss": 1.9568, "step": 2800 }, { "epoch": 0.92, "learning_rate": 7.785219433859847e-07, "loss": 2.0242, "step": 2810 }, { "epoch": 0.92, "learning_rate": 7.160658933218517e-07, "loss": 1.9992, "step": 2820 }, { "epoch": 0.93, "learning_rate": 6.561845664121297e-07, "loss": 2.0315, "step": 2830 }, { "epoch": 0.93, "learning_rate": 5.988843116214616e-07, "loss": 1.9966, "step": 2840 }, { "epoch": 0.93, "learning_rate": 5.441712042542845e-07, "loss": 1.9345, "step": 2850 }, { "epoch": 0.94, "learning_rate": 4.920510453107097e-07, "loss": 2.0039, "step": 2860 }, { "epoch": 0.94, "learning_rate": 4.425293608714548e-07, "loss": 1.9521, "step": 2870 }, { "epoch": 0.94, "learning_rate": 3.9561140151194166e-07, "loss": 1.9668, "step": 2880 }, { "epoch": 0.95, "learning_rate": 3.5130214174560193e-07, "loss": 2.0093, "step": 2890 }, { "epoch": 0.95, "learning_rate": 3.0960627949644107e-07, "loss": 1.991, "step": 2900 }, { "epoch": 0.95, "learning_rate": 2.705282356009503e-07, "loss": 1.9216, "step": 2910 }, { "epoch": 0.96, "learning_rate": 2.3407215333937082e-07, "loss": 1.8886, "step": 2920 }, { "epoch": 0.96, "learning_rate": 2.0024189799641435e-07, "loss": 2.0617, "step": 2930 }, { "epoch": 0.96, "learning_rate": 1.6904105645142444e-07, "loss": 2.106, "step": 2940 }, { "epoch": 0.97, "learning_rate": 1.4047293679809448e-07, "loss": 1.9519, "step": 2950 }, { "epoch": 0.97, "learning_rate": 1.1454056799370672e-07, "loss": 2.0065, "step": 2960 }, { "epoch": 0.97, "learning_rate": 9.124669953800014e-08, "loss": 1.9542, "step": 2970 }, { "epoch": 0.98, "learning_rate": 7.059380118163983e-08, "loss": 1.9162, "step": 2980 }, { "epoch": 0.98, "learning_rate": 5.2584062664368105e-08, "loss": 2.0596, "step": 2990 }, { "epoch": 0.98, "learning_rate": 3.7219393482834654e-08, "loss": 2.0616, "step": 3000 } ], "max_steps": 3051, "num_train_epochs": 1, "total_flos": 3.754066494874583e+17, "trial_name": null, "trial_params": null }