{ "best_metric": null, "best_model_checkpoint": null, "epoch": 9.992622895848703, "eval_steps": 500, "global_step": 149000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03, "learning_rate": 1.9932935416806386e-05, "loss": 0.7069, "step": 500 }, { "epoch": 0.07, "learning_rate": 1.986587083361277e-05, "loss": 0.4778, "step": 1000 }, { "epoch": 0.1, "learning_rate": 1.9798806250419156e-05, "loss": 0.4614, "step": 1500 }, { "epoch": 0.13, "learning_rate": 1.973174166722554e-05, "loss": 0.395, "step": 2000 }, { "epoch": 0.17, "learning_rate": 1.9664677084031925e-05, "loss": 0.3459, "step": 2500 }, { "epoch": 0.2, "learning_rate": 1.9597612500838306e-05, "loss": 0.3942, "step": 3000 }, { "epoch": 0.23, "learning_rate": 1.9530547917644694e-05, "loss": 0.5045, "step": 3500 }, { "epoch": 0.27, "learning_rate": 1.946348333445108e-05, "loss": 0.5222, "step": 4000 }, { "epoch": 0.3, "learning_rate": 1.9396418751257464e-05, "loss": 0.6322, "step": 4500 }, { "epoch": 0.34, "learning_rate": 1.9329354168063848e-05, "loss": 0.4139, "step": 5000 }, { "epoch": 0.37, "learning_rate": 1.926228958487023e-05, "loss": 0.3958, "step": 5500 }, { "epoch": 0.4, "learning_rate": 1.9195225001676618e-05, "loss": 0.3241, "step": 6000 }, { "epoch": 0.44, "learning_rate": 1.9128160418483002e-05, "loss": 0.3645, "step": 6500 }, { "epoch": 0.47, "learning_rate": 1.9061095835289383e-05, "loss": 0.4296, "step": 7000 }, { "epoch": 0.5, "learning_rate": 1.899403125209577e-05, "loss": 0.4467, "step": 7500 }, { "epoch": 0.54, "learning_rate": 1.8926966668902153e-05, "loss": 0.499, "step": 8000 }, { "epoch": 0.57, "learning_rate": 1.8859902085708537e-05, "loss": 0.3124, "step": 8500 }, { "epoch": 0.6, "learning_rate": 1.8792837502514922e-05, "loss": 0.4169, "step": 9000 }, { "epoch": 0.64, "learning_rate": 1.8725772919321307e-05, "loss": 0.4071, "step": 9500 }, { "epoch": 0.67, "learning_rate": 1.8658708336127695e-05, "loss": 0.2717, "step": 10000 }, { "epoch": 0.7, "learning_rate": 1.8591643752934076e-05, "loss": 0.3581, "step": 10500 }, { "epoch": 0.74, "learning_rate": 1.852457916974046e-05, "loss": 0.3181, "step": 11000 }, { "epoch": 0.77, "learning_rate": 1.8457514586546845e-05, "loss": 0.3488, "step": 11500 }, { "epoch": 0.8, "learning_rate": 1.839045000335323e-05, "loss": 0.3342, "step": 12000 }, { "epoch": 0.84, "learning_rate": 1.8323385420159615e-05, "loss": 0.4063, "step": 12500 }, { "epoch": 0.87, "learning_rate": 1.8256320836966e-05, "loss": 0.3071, "step": 13000 }, { "epoch": 0.91, "learning_rate": 1.8189256253772384e-05, "loss": 0.2998, "step": 13500 }, { "epoch": 0.94, "learning_rate": 1.812219167057877e-05, "loss": 0.2989, "step": 14000 }, { "epoch": 0.97, "learning_rate": 1.8055127087385153e-05, "loss": 0.2705, "step": 14500 }, { "epoch": 1.01, "learning_rate": 1.7988062504191538e-05, "loss": 0.3, "step": 15000 }, { "epoch": 1.04, "learning_rate": 1.7920997920997923e-05, "loss": 0.3253, "step": 15500 }, { "epoch": 1.07, "learning_rate": 1.7853933337804307e-05, "loss": 0.2612, "step": 16000 }, { "epoch": 1.11, "learning_rate": 1.7786868754610692e-05, "loss": 0.3027, "step": 16500 }, { "epoch": 1.14, "learning_rate": 1.7719804171417077e-05, "loss": 0.3124, "step": 17000 }, { "epoch": 1.17, "learning_rate": 1.765273958822346e-05, "loss": 0.2997, "step": 17500 }, { "epoch": 1.21, "learning_rate": 1.7585675005029846e-05, "loss": 0.3177, "step": 18000 }, { "epoch": 1.24, "learning_rate": 1.751861042183623e-05, "loss": 0.3636, "step": 18500 }, { "epoch": 1.27, "learning_rate": 1.7451545838642615e-05, "loss": 0.3025, "step": 19000 }, { "epoch": 1.31, "learning_rate": 1.7384481255449e-05, "loss": 0.4098, "step": 19500 }, { "epoch": 1.34, "learning_rate": 1.731741667225538e-05, "loss": 0.3093, "step": 20000 }, { "epoch": 1.37, "learning_rate": 1.725035208906177e-05, "loss": 0.3261, "step": 20500 }, { "epoch": 1.41, "learning_rate": 1.718328750586815e-05, "loss": 0.2338, "step": 21000 }, { "epoch": 1.44, "learning_rate": 1.711622292267454e-05, "loss": 0.2204, "step": 21500 }, { "epoch": 1.48, "learning_rate": 1.7049158339480923e-05, "loss": 0.2963, "step": 22000 }, { "epoch": 1.51, "learning_rate": 1.6982093756287304e-05, "loss": 0.1368, "step": 22500 }, { "epoch": 1.54, "learning_rate": 1.6915029173093693e-05, "loss": 0.307, "step": 23000 }, { "epoch": 1.58, "learning_rate": 1.6847964589900074e-05, "loss": 0.2184, "step": 23500 }, { "epoch": 1.61, "learning_rate": 1.678090000670646e-05, "loss": 0.2967, "step": 24000 }, { "epoch": 1.64, "learning_rate": 1.6713835423512847e-05, "loss": 0.2879, "step": 24500 }, { "epoch": 1.68, "learning_rate": 1.6646770840319228e-05, "loss": 0.2899, "step": 25000 }, { "epoch": 1.71, "learning_rate": 1.6579706257125612e-05, "loss": 0.2182, "step": 25500 }, { "epoch": 1.74, "learning_rate": 1.6512641673931997e-05, "loss": 0.1737, "step": 26000 }, { "epoch": 1.78, "learning_rate": 1.6445577090738382e-05, "loss": 0.2756, "step": 26500 }, { "epoch": 1.81, "learning_rate": 1.6378512507544766e-05, "loss": 0.284, "step": 27000 }, { "epoch": 1.84, "learning_rate": 1.631144792435115e-05, "loss": 0.2672, "step": 27500 }, { "epoch": 1.88, "learning_rate": 1.6244383341157536e-05, "loss": 0.2016, "step": 28000 }, { "epoch": 1.91, "learning_rate": 1.617731875796392e-05, "loss": 0.2257, "step": 28500 }, { "epoch": 1.94, "learning_rate": 1.6110254174770305e-05, "loss": 0.3434, "step": 29000 }, { "epoch": 1.98, "learning_rate": 1.604318959157669e-05, "loss": 0.1699, "step": 29500 }, { "epoch": 2.01, "learning_rate": 1.5976125008383074e-05, "loss": 0.2511, "step": 30000 }, { "epoch": 2.05, "learning_rate": 1.590906042518946e-05, "loss": 0.2026, "step": 30500 }, { "epoch": 2.08, "learning_rate": 1.5841995841995844e-05, "loss": 0.2522, "step": 31000 }, { "epoch": 2.11, "learning_rate": 1.577493125880223e-05, "loss": 0.2572, "step": 31500 }, { "epoch": 2.15, "learning_rate": 1.5707866675608613e-05, "loss": 0.2237, "step": 32000 }, { "epoch": 2.18, "learning_rate": 1.5640802092414998e-05, "loss": 0.195, "step": 32500 }, { "epoch": 2.21, "learning_rate": 1.557373750922138e-05, "loss": 0.218, "step": 33000 }, { "epoch": 2.25, "learning_rate": 1.5506672926027767e-05, "loss": 0.1463, "step": 33500 }, { "epoch": 2.28, "learning_rate": 1.543960834283415e-05, "loss": 0.1723, "step": 34000 }, { "epoch": 2.31, "learning_rate": 1.5372543759640536e-05, "loss": 0.1874, "step": 34500 }, { "epoch": 2.35, "learning_rate": 1.530547917644692e-05, "loss": 0.2871, "step": 35000 }, { "epoch": 2.38, "learning_rate": 1.5238414593253304e-05, "loss": 0.3132, "step": 35500 }, { "epoch": 2.41, "learning_rate": 1.5171350010059689e-05, "loss": 0.2591, "step": 36000 }, { "epoch": 2.45, "learning_rate": 1.5104285426866075e-05, "loss": 0.1916, "step": 36500 }, { "epoch": 2.48, "learning_rate": 1.5037220843672458e-05, "loss": 0.2998, "step": 37000 }, { "epoch": 2.51, "learning_rate": 1.4970156260478843e-05, "loss": 0.1397, "step": 37500 }, { "epoch": 2.55, "learning_rate": 1.4903091677285225e-05, "loss": 0.2588, "step": 38000 }, { "epoch": 2.58, "learning_rate": 1.4836027094091612e-05, "loss": 0.2526, "step": 38500 }, { "epoch": 2.62, "learning_rate": 1.4768962510897995e-05, "loss": 0.1548, "step": 39000 }, { "epoch": 2.65, "learning_rate": 1.4701897927704381e-05, "loss": 0.2322, "step": 39500 }, { "epoch": 2.68, "learning_rate": 1.4634833344510766e-05, "loss": 0.1839, "step": 40000 }, { "epoch": 2.72, "learning_rate": 1.4567768761317149e-05, "loss": 0.1062, "step": 40500 }, { "epoch": 2.75, "learning_rate": 1.4500704178123535e-05, "loss": 0.1758, "step": 41000 }, { "epoch": 2.78, "learning_rate": 1.4433639594929918e-05, "loss": 0.2261, "step": 41500 }, { "epoch": 2.82, "learning_rate": 1.4366575011736303e-05, "loss": 0.1591, "step": 42000 }, { "epoch": 2.85, "learning_rate": 1.4299510428542687e-05, "loss": 0.2023, "step": 42500 }, { "epoch": 2.88, "learning_rate": 1.4232445845349072e-05, "loss": 0.2378, "step": 43000 }, { "epoch": 2.92, "learning_rate": 1.4165381262155458e-05, "loss": 0.1869, "step": 43500 }, { "epoch": 2.95, "learning_rate": 1.4098316678961841e-05, "loss": 0.2229, "step": 44000 }, { "epoch": 2.98, "learning_rate": 1.4031252095768226e-05, "loss": 0.1752, "step": 44500 }, { "epoch": 3.02, "learning_rate": 1.3964187512574609e-05, "loss": 0.2121, "step": 45000 }, { "epoch": 3.05, "learning_rate": 1.3897122929380995e-05, "loss": 0.1928, "step": 45500 }, { "epoch": 3.08, "learning_rate": 1.383005834618738e-05, "loss": 0.0827, "step": 46000 }, { "epoch": 3.12, "learning_rate": 1.3762993762993763e-05, "loss": 0.2259, "step": 46500 }, { "epoch": 3.15, "learning_rate": 1.369592917980015e-05, "loss": 0.1447, "step": 47000 }, { "epoch": 3.19, "learning_rate": 1.3628864596606532e-05, "loss": 0.1481, "step": 47500 }, { "epoch": 3.22, "learning_rate": 1.3561800013412919e-05, "loss": 0.2343, "step": 48000 }, { "epoch": 3.25, "learning_rate": 1.3494735430219302e-05, "loss": 0.1454, "step": 48500 }, { "epoch": 3.29, "learning_rate": 1.3427670847025686e-05, "loss": 0.1152, "step": 49000 }, { "epoch": 3.32, "learning_rate": 1.3360606263832073e-05, "loss": 0.2361, "step": 49500 }, { "epoch": 3.35, "learning_rate": 1.3293541680638456e-05, "loss": 0.091, "step": 50000 }, { "epoch": 3.39, "learning_rate": 1.322647709744484e-05, "loss": 0.1912, "step": 50500 }, { "epoch": 3.42, "learning_rate": 1.3159412514251225e-05, "loss": 0.1125, "step": 51000 }, { "epoch": 3.45, "learning_rate": 1.309234793105761e-05, "loss": 0.1295, "step": 51500 }, { "epoch": 3.49, "learning_rate": 1.3025283347863994e-05, "loss": 0.106, "step": 52000 }, { "epoch": 3.52, "learning_rate": 1.2958218764670379e-05, "loss": 0.0838, "step": 52500 }, { "epoch": 3.55, "learning_rate": 1.2891154181476764e-05, "loss": 0.1616, "step": 53000 }, { "epoch": 3.59, "learning_rate": 1.2824089598283147e-05, "loss": 0.1084, "step": 53500 }, { "epoch": 3.62, "learning_rate": 1.2757025015089533e-05, "loss": 0.2535, "step": 54000 }, { "epoch": 3.66, "learning_rate": 1.2689960431895916e-05, "loss": 0.1363, "step": 54500 }, { "epoch": 3.69, "learning_rate": 1.26228958487023e-05, "loss": 0.2317, "step": 55000 }, { "epoch": 3.72, "learning_rate": 1.2555831265508687e-05, "loss": 0.2266, "step": 55500 }, { "epoch": 3.76, "learning_rate": 1.248876668231507e-05, "loss": 0.1864, "step": 56000 }, { "epoch": 3.79, "learning_rate": 1.2421702099121456e-05, "loss": 0.0946, "step": 56500 }, { "epoch": 3.82, "learning_rate": 1.2354637515927839e-05, "loss": 0.1919, "step": 57000 }, { "epoch": 3.86, "learning_rate": 1.2287572932734224e-05, "loss": 0.1106, "step": 57500 }, { "epoch": 3.89, "learning_rate": 1.222050834954061e-05, "loss": 0.1769, "step": 58000 }, { "epoch": 3.92, "learning_rate": 1.2153443766346993e-05, "loss": 0.089, "step": 58500 }, { "epoch": 3.96, "learning_rate": 1.2086379183153378e-05, "loss": 0.1087, "step": 59000 }, { "epoch": 3.99, "learning_rate": 1.2019314599959762e-05, "loss": 0.1879, "step": 59500 }, { "epoch": 4.02, "learning_rate": 1.1952250016766147e-05, "loss": 0.14, "step": 60000 }, { "epoch": 4.06, "learning_rate": 1.188518543357253e-05, "loss": 0.1331, "step": 60500 }, { "epoch": 4.09, "learning_rate": 1.1818120850378916e-05, "loss": 0.0841, "step": 61000 }, { "epoch": 4.12, "learning_rate": 1.1751056267185301e-05, "loss": 0.134, "step": 61500 }, { "epoch": 4.16, "learning_rate": 1.1683991683991684e-05, "loss": 0.1823, "step": 62000 }, { "epoch": 4.19, "learning_rate": 1.161692710079807e-05, "loss": 0.1005, "step": 62500 }, { "epoch": 4.23, "learning_rate": 1.1549862517604453e-05, "loss": 0.1003, "step": 63000 }, { "epoch": 4.26, "learning_rate": 1.1482797934410838e-05, "loss": 0.1729, "step": 63500 }, { "epoch": 4.29, "learning_rate": 1.1415733351217224e-05, "loss": 0.155, "step": 64000 }, { "epoch": 4.33, "learning_rate": 1.1348668768023607e-05, "loss": 0.1039, "step": 64500 }, { "epoch": 4.36, "learning_rate": 1.1281604184829994e-05, "loss": 0.1536, "step": 65000 }, { "epoch": 4.39, "learning_rate": 1.1214539601636377e-05, "loss": 0.1576, "step": 65500 }, { "epoch": 4.43, "learning_rate": 1.1147475018442761e-05, "loss": 0.1298, "step": 66000 }, { "epoch": 4.46, "learning_rate": 1.1080410435249144e-05, "loss": 0.1465, "step": 66500 }, { "epoch": 4.49, "learning_rate": 1.101334585205553e-05, "loss": 0.1322, "step": 67000 }, { "epoch": 4.53, "learning_rate": 1.0946281268861915e-05, "loss": 0.1322, "step": 67500 }, { "epoch": 4.56, "learning_rate": 1.0879216685668298e-05, "loss": 0.1597, "step": 68000 }, { "epoch": 4.59, "learning_rate": 1.0812152102474685e-05, "loss": 0.1188, "step": 68500 }, { "epoch": 4.63, "learning_rate": 1.0745087519281068e-05, "loss": 0.1614, "step": 69000 }, { "epoch": 4.66, "learning_rate": 1.0678022936087454e-05, "loss": 0.1174, "step": 69500 }, { "epoch": 4.69, "learning_rate": 1.0610958352893839e-05, "loss": 0.1336, "step": 70000 }, { "epoch": 4.73, "learning_rate": 1.0543893769700222e-05, "loss": 0.0941, "step": 70500 }, { "epoch": 4.76, "learning_rate": 1.0476829186506608e-05, "loss": 0.0724, "step": 71000 }, { "epoch": 4.8, "learning_rate": 1.040976460331299e-05, "loss": 0.1419, "step": 71500 }, { "epoch": 4.83, "learning_rate": 1.0342700020119375e-05, "loss": 0.1308, "step": 72000 }, { "epoch": 4.86, "learning_rate": 1.027563543692576e-05, "loss": 0.1128, "step": 72500 }, { "epoch": 4.9, "learning_rate": 1.0208570853732145e-05, "loss": 0.0805, "step": 73000 }, { "epoch": 4.93, "learning_rate": 1.0141506270538531e-05, "loss": 0.0361, "step": 73500 }, { "epoch": 4.96, "learning_rate": 1.0074441687344914e-05, "loss": 0.1846, "step": 74000 }, { "epoch": 5.0, "learning_rate": 1.0007377104151299e-05, "loss": 0.0949, "step": 74500 }, { "epoch": 5.03, "learning_rate": 9.940312520957683e-06, "loss": 0.1941, "step": 75000 }, { "epoch": 5.06, "learning_rate": 9.873247937764068e-06, "loss": 0.154, "step": 75500 }, { "epoch": 5.1, "learning_rate": 9.806183354570453e-06, "loss": 0.1002, "step": 76000 }, { "epoch": 5.13, "learning_rate": 9.739118771376836e-06, "loss": 0.0869, "step": 76500 }, { "epoch": 5.16, "learning_rate": 9.67205418818322e-06, "loss": 0.0587, "step": 77000 }, { "epoch": 5.2, "learning_rate": 9.604989604989607e-06, "loss": 0.1265, "step": 77500 }, { "epoch": 5.23, "learning_rate": 9.537925021795991e-06, "loss": 0.1247, "step": 78000 }, { "epoch": 5.26, "learning_rate": 9.470860438602374e-06, "loss": 0.0767, "step": 78500 }, { "epoch": 5.3, "learning_rate": 9.403795855408759e-06, "loss": 0.0519, "step": 79000 }, { "epoch": 5.33, "learning_rate": 9.336731272215144e-06, "loss": 0.0949, "step": 79500 }, { "epoch": 5.37, "learning_rate": 9.269666689021528e-06, "loss": 0.085, "step": 80000 }, { "epoch": 5.4, "learning_rate": 9.202602105827913e-06, "loss": 0.1271, "step": 80500 }, { "epoch": 5.43, "learning_rate": 9.135537522634298e-06, "loss": 0.2085, "step": 81000 }, { "epoch": 5.47, "learning_rate": 9.068472939440682e-06, "loss": 0.0912, "step": 81500 }, { "epoch": 5.5, "learning_rate": 9.001408356247067e-06, "loss": 0.1413, "step": 82000 }, { "epoch": 5.53, "learning_rate": 8.934343773053452e-06, "loss": 0.1015, "step": 82500 }, { "epoch": 5.57, "learning_rate": 8.867279189859835e-06, "loss": 0.1277, "step": 83000 }, { "epoch": 5.6, "learning_rate": 8.80021460666622e-06, "loss": 0.0239, "step": 83500 }, { "epoch": 5.63, "learning_rate": 8.733150023472606e-06, "loss": 0.1947, "step": 84000 }, { "epoch": 5.67, "learning_rate": 8.66608544027899e-06, "loss": 0.0589, "step": 84500 }, { "epoch": 5.7, "learning_rate": 8.599020857085373e-06, "loss": 0.112, "step": 85000 }, { "epoch": 5.73, "learning_rate": 8.531956273891758e-06, "loss": 0.0843, "step": 85500 }, { "epoch": 5.77, "learning_rate": 8.464891690698143e-06, "loss": 0.151, "step": 86000 }, { "epoch": 5.8, "learning_rate": 8.397827107504527e-06, "loss": 0.1011, "step": 86500 }, { "epoch": 5.83, "learning_rate": 8.330762524310912e-06, "loss": 0.045, "step": 87000 }, { "epoch": 5.87, "learning_rate": 8.263697941117297e-06, "loss": 0.0533, "step": 87500 }, { "epoch": 5.9, "learning_rate": 8.196633357923681e-06, "loss": 0.0512, "step": 88000 }, { "epoch": 5.94, "learning_rate": 8.129568774730066e-06, "loss": 0.2159, "step": 88500 }, { "epoch": 5.97, "learning_rate": 8.06250419153645e-06, "loss": 0.1077, "step": 89000 }, { "epoch": 6.0, "learning_rate": 7.995439608342833e-06, "loss": 0.11, "step": 89500 }, { "epoch": 6.04, "learning_rate": 7.92837502514922e-06, "loss": 0.1365, "step": 90000 }, { "epoch": 6.07, "learning_rate": 7.861310441955604e-06, "loss": 0.0742, "step": 90500 }, { "epoch": 6.1, "learning_rate": 7.794245858761989e-06, "loss": 0.0716, "step": 91000 }, { "epoch": 6.14, "learning_rate": 7.727181275568372e-06, "loss": 0.1481, "step": 91500 }, { "epoch": 6.17, "learning_rate": 7.660116692374757e-06, "loss": 0.008, "step": 92000 }, { "epoch": 6.2, "learning_rate": 7.593052109181141e-06, "loss": 0.0615, "step": 92500 }, { "epoch": 6.24, "learning_rate": 7.525987525987527e-06, "loss": 0.106, "step": 93000 }, { "epoch": 6.27, "learning_rate": 7.4589229427939116e-06, "loss": 0.0534, "step": 93500 }, { "epoch": 6.3, "learning_rate": 7.391858359600295e-06, "loss": 0.1034, "step": 94000 }, { "epoch": 6.34, "learning_rate": 7.32479377640668e-06, "loss": 0.107, "step": 94500 }, { "epoch": 6.37, "learning_rate": 7.257729193213065e-06, "loss": 0.0235, "step": 95000 }, { "epoch": 6.4, "learning_rate": 7.1906646100194485e-06, "loss": 0.0844, "step": 95500 }, { "epoch": 6.44, "learning_rate": 7.123600026825834e-06, "loss": 0.0495, "step": 96000 }, { "epoch": 6.47, "learning_rate": 7.056535443632219e-06, "loss": 0.0572, "step": 96500 }, { "epoch": 6.51, "learning_rate": 6.989470860438603e-06, "loss": 0.024, "step": 97000 }, { "epoch": 6.54, "learning_rate": 6.922406277244987e-06, "loss": 0.1605, "step": 97500 }, { "epoch": 6.57, "learning_rate": 6.855341694051372e-06, "loss": 0.1724, "step": 98000 }, { "epoch": 6.61, "learning_rate": 6.7882771108577564e-06, "loss": 0.0505, "step": 98500 }, { "epoch": 6.64, "learning_rate": 6.721212527664142e-06, "loss": 0.0549, "step": 99000 }, { "epoch": 6.67, "learning_rate": 6.654147944470526e-06, "loss": 0.1133, "step": 99500 }, { "epoch": 6.71, "learning_rate": 6.58708336127691e-06, "loss": 0.0441, "step": 100000 }, { "epoch": 6.74, "learning_rate": 6.520018778083294e-06, "loss": 0.0775, "step": 100500 }, { "epoch": 6.77, "learning_rate": 6.452954194889679e-06, "loss": 0.0319, "step": 101000 }, { "epoch": 6.81, "learning_rate": 6.3858896116960635e-06, "loss": 0.1535, "step": 101500 }, { "epoch": 6.84, "learning_rate": 6.318825028502449e-06, "loss": 0.0782, "step": 102000 }, { "epoch": 6.87, "learning_rate": 6.251760445308833e-06, "loss": 0.1007, "step": 102500 }, { "epoch": 6.91, "learning_rate": 6.1846958621152175e-06, "loss": 0.0677, "step": 103000 }, { "epoch": 6.94, "learning_rate": 6.117631278921602e-06, "loss": 0.0279, "step": 103500 }, { "epoch": 6.97, "learning_rate": 6.050566695727986e-06, "loss": 0.0685, "step": 104000 }, { "epoch": 7.01, "learning_rate": 5.983502112534371e-06, "loss": 0.0705, "step": 104500 }, { "epoch": 7.04, "learning_rate": 5.916437529340756e-06, "loss": 0.0993, "step": 105000 }, { "epoch": 7.08, "learning_rate": 5.849372946147141e-06, "loss": 0.0558, "step": 105500 }, { "epoch": 7.11, "learning_rate": 5.782308362953525e-06, "loss": 0.0672, "step": 106000 }, { "epoch": 7.14, "learning_rate": 5.715243779759909e-06, "loss": 0.0796, "step": 106500 }, { "epoch": 7.18, "learning_rate": 5.648179196566294e-06, "loss": 0.038, "step": 107000 }, { "epoch": 7.21, "learning_rate": 5.581114613372678e-06, "loss": 0.0612, "step": 107500 }, { "epoch": 7.24, "learning_rate": 5.514050030179063e-06, "loss": 0.0537, "step": 108000 }, { "epoch": 7.28, "learning_rate": 5.446985446985448e-06, "loss": 0.0291, "step": 108500 }, { "epoch": 7.31, "learning_rate": 5.379920863791832e-06, "loss": 0.0772, "step": 109000 }, { "epoch": 7.34, "learning_rate": 5.312856280598216e-06, "loss": 0.0997, "step": 109500 }, { "epoch": 7.38, "learning_rate": 5.245791697404601e-06, "loss": 0.0505, "step": 110000 }, { "epoch": 7.41, "learning_rate": 5.178727114210985e-06, "loss": 0.0636, "step": 110500 }, { "epoch": 7.44, "learning_rate": 5.11166253101737e-06, "loss": 0.0314, "step": 111000 }, { "epoch": 7.48, "learning_rate": 5.044597947823755e-06, "loss": 0.1147, "step": 111500 }, { "epoch": 7.51, "learning_rate": 4.97753336463014e-06, "loss": 0.0815, "step": 112000 }, { "epoch": 7.54, "learning_rate": 4.9104687814365235e-06, "loss": 0.016, "step": 112500 }, { "epoch": 7.58, "learning_rate": 4.843404198242908e-06, "loss": 0.0509, "step": 113000 }, { "epoch": 7.61, "learning_rate": 4.776339615049293e-06, "loss": 0.0365, "step": 113500 }, { "epoch": 7.65, "learning_rate": 4.7092750318556775e-06, "loss": 0.0374, "step": 114000 }, { "epoch": 7.68, "learning_rate": 4.642210448662062e-06, "loss": 0.1001, "step": 114500 }, { "epoch": 7.71, "learning_rate": 4.575145865468447e-06, "loss": 0.021, "step": 115000 }, { "epoch": 7.75, "learning_rate": 4.508081282274831e-06, "loss": 0.0171, "step": 115500 }, { "epoch": 7.78, "learning_rate": 4.441016699081215e-06, "loss": 0.0759, "step": 116000 }, { "epoch": 7.81, "learning_rate": 4.3739521158876e-06, "loss": 0.0804, "step": 116500 }, { "epoch": 7.85, "learning_rate": 4.3068875326939846e-06, "loss": 0.0418, "step": 117000 }, { "epoch": 7.88, "learning_rate": 4.239822949500369e-06, "loss": 0.1122, "step": 117500 }, { "epoch": 7.91, "learning_rate": 4.172758366306754e-06, "loss": 0.0737, "step": 118000 }, { "epoch": 7.95, "learning_rate": 4.1056937831131385e-06, "loss": 0.0526, "step": 118500 }, { "epoch": 7.98, "learning_rate": 4.038629199919522e-06, "loss": 0.0483, "step": 119000 }, { "epoch": 8.01, "learning_rate": 3.971564616725908e-06, "loss": 0.037, "step": 119500 }, { "epoch": 8.05, "learning_rate": 3.904500033532292e-06, "loss": 0.0687, "step": 120000 }, { "epoch": 8.08, "learning_rate": 3.837435450338676e-06, "loss": 0.023, "step": 120500 }, { "epoch": 8.11, "learning_rate": 3.770370867145061e-06, "loss": 0.1, "step": 121000 }, { "epoch": 8.15, "learning_rate": 3.7033062839514456e-06, "loss": 0.0659, "step": 121500 }, { "epoch": 8.18, "learning_rate": 3.63624170075783e-06, "loss": 0.0989, "step": 122000 }, { "epoch": 8.22, "learning_rate": 3.569177117564215e-06, "loss": 0.0603, "step": 122500 }, { "epoch": 8.25, "learning_rate": 3.502112534370599e-06, "loss": 0.0321, "step": 123000 }, { "epoch": 8.28, "learning_rate": 3.4350479511769834e-06, "loss": 0.0709, "step": 123500 }, { "epoch": 8.32, "learning_rate": 3.3679833679833685e-06, "loss": 0.0389, "step": 124000 }, { "epoch": 8.35, "learning_rate": 3.3009187847897527e-06, "loss": 0.036, "step": 124500 }, { "epoch": 8.38, "learning_rate": 3.233854201596137e-06, "loss": 0.04, "step": 125000 }, { "epoch": 8.42, "learning_rate": 3.166789618402522e-06, "loss": 0.0265, "step": 125500 }, { "epoch": 8.45, "learning_rate": 3.0997250352089063e-06, "loss": 0.0434, "step": 126000 }, { "epoch": 8.48, "learning_rate": 3.032660452015291e-06, "loss": 0.0001, "step": 126500 }, { "epoch": 8.52, "learning_rate": 2.9655958688216756e-06, "loss": 0.0224, "step": 127000 }, { "epoch": 8.55, "learning_rate": 2.8985312856280603e-06, "loss": 0.0199, "step": 127500 }, { "epoch": 8.58, "learning_rate": 2.8314667024344445e-06, "loss": 0.0482, "step": 128000 }, { "epoch": 8.62, "learning_rate": 2.764402119240829e-06, "loss": 0.0329, "step": 128500 }, { "epoch": 8.65, "learning_rate": 2.697337536047214e-06, "loss": 0.0515, "step": 129000 }, { "epoch": 8.68, "learning_rate": 2.630272952853598e-06, "loss": 0.0473, "step": 129500 }, { "epoch": 8.72, "learning_rate": 2.563208369659983e-06, "loss": 0.1201, "step": 130000 }, { "epoch": 8.75, "learning_rate": 2.4961437864663674e-06, "loss": 0.0295, "step": 130500 }, { "epoch": 8.79, "learning_rate": 2.429079203272752e-06, "loss": 0.0001, "step": 131000 }, { "epoch": 8.82, "learning_rate": 2.3620146200791363e-06, "loss": 0.0537, "step": 131500 }, { "epoch": 8.85, "learning_rate": 2.294950036885521e-06, "loss": 0.0821, "step": 132000 }, { "epoch": 8.89, "learning_rate": 2.2278854536919056e-06, "loss": 0.0665, "step": 132500 }, { "epoch": 8.92, "learning_rate": 2.16082087049829e-06, "loss": 0.0391, "step": 133000 }, { "epoch": 8.95, "learning_rate": 2.0937562873046745e-06, "loss": 0.0315, "step": 133500 }, { "epoch": 8.99, "learning_rate": 2.026691704111059e-06, "loss": 0.0431, "step": 134000 }, { "epoch": 9.02, "learning_rate": 1.959627120917444e-06, "loss": 0.0328, "step": 134500 }, { "epoch": 9.05, "learning_rate": 1.8925625377238282e-06, "loss": 0.0126, "step": 135000 }, { "epoch": 9.09, "learning_rate": 1.8254979545302127e-06, "loss": 0.0086, "step": 135500 }, { "epoch": 9.12, "learning_rate": 1.7584333713365973e-06, "loss": 0.0242, "step": 136000 }, { "epoch": 9.15, "learning_rate": 1.6913687881429818e-06, "loss": 0.0983, "step": 136500 }, { "epoch": 9.19, "learning_rate": 1.6243042049493662e-06, "loss": 0.0471, "step": 137000 }, { "epoch": 9.22, "learning_rate": 1.557239621755751e-06, "loss": 0.0347, "step": 137500 }, { "epoch": 9.25, "learning_rate": 1.4901750385621356e-06, "loss": 0.0387, "step": 138000 }, { "epoch": 9.29, "learning_rate": 1.4231104553685198e-06, "loss": 0.0398, "step": 138500 }, { "epoch": 9.32, "learning_rate": 1.3560458721749044e-06, "loss": 0.062, "step": 139000 }, { "epoch": 9.36, "learning_rate": 1.288981288981289e-06, "loss": 0.0003, "step": 139500 }, { "epoch": 9.39, "learning_rate": 1.2219167057876738e-06, "loss": 0.0466, "step": 140000 }, { "epoch": 9.42, "learning_rate": 1.1548521225940582e-06, "loss": 0.0385, "step": 140500 }, { "epoch": 9.46, "learning_rate": 1.0877875394004427e-06, "loss": 0.0581, "step": 141000 }, { "epoch": 9.49, "learning_rate": 1.0207229562068273e-06, "loss": 0.016, "step": 141500 }, { "epoch": 9.52, "learning_rate": 9.536583730132118e-07, "loss": 0.0202, "step": 142000 }, { "epoch": 9.56, "learning_rate": 8.865937898195963e-07, "loss": 0.0409, "step": 142500 }, { "epoch": 9.59, "learning_rate": 8.19529206625981e-07, "loss": 0.0327, "step": 143000 }, { "epoch": 9.62, "learning_rate": 7.524646234323654e-07, "loss": 0.0101, "step": 143500 }, { "epoch": 9.66, "learning_rate": 6.8540004023875e-07, "loss": 0.0539, "step": 144000 }, { "epoch": 9.69, "learning_rate": 6.183354570451345e-07, "loss": 0.0001, "step": 144500 }, { "epoch": 9.72, "learning_rate": 5.512708738515191e-07, "loss": 0.0521, "step": 145000 }, { "epoch": 9.76, "learning_rate": 4.842062906579036e-07, "loss": 0.0702, "step": 145500 }, { "epoch": 9.79, "learning_rate": 4.1714170746428813e-07, "loss": 0.0559, "step": 146000 }, { "epoch": 9.82, "learning_rate": 3.500771242706727e-07, "loss": 0.0394, "step": 146500 }, { "epoch": 9.86, "learning_rate": 2.8301254107705723e-07, "loss": 0.0164, "step": 147000 }, { "epoch": 9.89, "learning_rate": 2.1594795788344176e-07, "loss": 0.0194, "step": 147500 }, { "epoch": 9.93, "learning_rate": 1.488833746898263e-07, "loss": 0.0652, "step": 148000 }, { "epoch": 9.96, "learning_rate": 8.181879149621086e-08, "loss": 0.077, "step": 148500 }, { "epoch": 9.99, "learning_rate": 1.4754208302595401e-08, "loss": 0.0659, "step": 149000 } ], "logging_steps": 500, "max_steps": 149110, "num_train_epochs": 10, "save_steps": 500, "total_flos": 9801062808576000.0, "trial_name": null, "trial_params": null }