| { | |
| "best_metric": 0.34630635380744934, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 2.197080291970803, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014598540145985401, | |
| "grad_norm": 33.662986755371094, | |
| "learning_rate": 1e-05, | |
| "loss": 6.8288, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.014598540145985401, | |
| "eval_loss": 6.846549987792969, | |
| "eval_runtime": 8.1835, | |
| "eval_samples_per_second": 14.053, | |
| "eval_steps_per_second": 3.544, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.029197080291970802, | |
| "grad_norm": 38.6097526550293, | |
| "learning_rate": 2e-05, | |
| "loss": 6.8553, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.043795620437956206, | |
| "grad_norm": 39.742008209228516, | |
| "learning_rate": 3e-05, | |
| "loss": 6.6102, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.058394160583941604, | |
| "grad_norm": 35.009727478027344, | |
| "learning_rate": 4e-05, | |
| "loss": 5.16, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.072992700729927, | |
| "grad_norm": 21.071718215942383, | |
| "learning_rate": 5e-05, | |
| "loss": 2.973, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08759124087591241, | |
| "grad_norm": 11.949044227600098, | |
| "learning_rate": 6e-05, | |
| "loss": 1.5887, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.10218978102189781, | |
| "grad_norm": 3.4320690631866455, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8089, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.11678832116788321, | |
| "grad_norm": 4.1433844566345215, | |
| "learning_rate": 8e-05, | |
| "loss": 0.6951, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.13138686131386862, | |
| "grad_norm": 2.0591485500335693, | |
| "learning_rate": 9e-05, | |
| "loss": 0.6693, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.145985401459854, | |
| "grad_norm": 0.7576547265052795, | |
| "learning_rate": 0.0001, | |
| "loss": 0.6251, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16058394160583941, | |
| "grad_norm": 8.802155494689941, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 1.0039, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.17518248175182483, | |
| "grad_norm": 2.171154737472534, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 0.7494, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1897810218978102, | |
| "grad_norm": 34.5106201171875, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 0.9494, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.20437956204379562, | |
| "grad_norm": 0.998836874961853, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 0.6674, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.21897810218978103, | |
| "grad_norm": 0.34883710741996765, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 0.6382, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.23357664233576642, | |
| "grad_norm": 2.8341825008392334, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 0.7636, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.24817518248175183, | |
| "grad_norm": 1.3019286394119263, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 0.7001, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.26277372262773724, | |
| "grad_norm": 0.6097630262374878, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 0.6695, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2773722627737226, | |
| "grad_norm": 0.5195863842964172, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 0.6628, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.291970802919708, | |
| "grad_norm": 1.1216007471084595, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 0.6775, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.30656934306569344, | |
| "grad_norm": 1.6819016933441162, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 0.7628, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.32116788321167883, | |
| "grad_norm": 0.5767019391059875, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 0.6679, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3357664233576642, | |
| "grad_norm": 0.6141683459281921, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 0.6266, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.35036496350364965, | |
| "grad_norm": 0.6812053918838501, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 0.5757, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.36496350364963503, | |
| "grad_norm": 3.8009650707244873, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 0.7547, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3795620437956204, | |
| "grad_norm": 5.604429244995117, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 1.0209, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.39416058394160586, | |
| "grad_norm": 2.7158730030059814, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 0.7527, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.40875912408759124, | |
| "grad_norm": 1.0342605113983154, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 0.7118, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4233576642335766, | |
| "grad_norm": 0.5667628049850464, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.6418, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.43795620437956206, | |
| "grad_norm": 0.42152413725852966, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 0.6735, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.45255474452554745, | |
| "grad_norm": 1.0017098188400269, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 0.7253, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.46715328467153283, | |
| "grad_norm": 0.56270432472229, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 0.6713, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.48175182481751827, | |
| "grad_norm": 0.8998551368713379, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 0.6585, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.49635036496350365, | |
| "grad_norm": 0.5632476806640625, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 0.7139, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5109489051094891, | |
| "grad_norm": 0.6208569407463074, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 0.6542, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5255474452554745, | |
| "grad_norm": 0.9806542992591858, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 0.674, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5401459854014599, | |
| "grad_norm": 0.3306947946548462, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 0.6498, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.5547445255474452, | |
| "grad_norm": 0.32137709856033325, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 0.5396, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5693430656934306, | |
| "grad_norm": 0.6273882985115051, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 0.6116, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.583941605839416, | |
| "grad_norm": 1.3096363544464111, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 0.7802, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5985401459854015, | |
| "grad_norm": 0.6243460774421692, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 0.6692, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6131386861313869, | |
| "grad_norm": 0.8591530323028564, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 0.725, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6277372262773723, | |
| "grad_norm": 0.5638594627380371, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 0.6857, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6423357664233577, | |
| "grad_norm": 0.8377383947372437, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 0.6591, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.656934306569343, | |
| "grad_norm": 0.6511130928993225, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 0.7041, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6715328467153284, | |
| "grad_norm": 0.9163225889205933, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 0.6581, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6861313868613139, | |
| "grad_norm": 0.71990567445755, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 0.6998, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7007299270072993, | |
| "grad_norm": 0.4144147038459778, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.7007, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7153284671532847, | |
| "grad_norm": 0.35297536849975586, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 0.6089, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7299270072992701, | |
| "grad_norm": 0.28604188561439514, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 0.674, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7299270072992701, | |
| "eval_loss": 0.6620756983757019, | |
| "eval_runtime": 8.3706, | |
| "eval_samples_per_second": 13.739, | |
| "eval_steps_per_second": 3.465, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7445255474452555, | |
| "grad_norm": 0.6144942045211792, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 0.6887, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.7591240875912408, | |
| "grad_norm": 0.44584447145462036, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 0.6056, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.7737226277372263, | |
| "grad_norm": 0.5202380418777466, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 0.6939, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.7883211678832117, | |
| "grad_norm": 0.3601526618003845, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 0.6728, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8029197080291971, | |
| "grad_norm": 0.42665866017341614, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 0.6846, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8175182481751825, | |
| "grad_norm": 0.630306601524353, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 0.6996, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8321167883211679, | |
| "grad_norm": 0.48954012989997864, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 0.6804, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.8467153284671532, | |
| "grad_norm": 0.797709047794342, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 0.6244, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.8613138686131386, | |
| "grad_norm": 0.470983624458313, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 0.6413, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.8759124087591241, | |
| "grad_norm": 0.5941108465194702, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 0.6347, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8905109489051095, | |
| "grad_norm": 0.32307079434394836, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 0.6821, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9051094890510949, | |
| "grad_norm": 0.4610147178173065, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 0.6431, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.9197080291970803, | |
| "grad_norm": 0.6533463597297668, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 0.6458, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.9343065693430657, | |
| "grad_norm": 0.2629701495170593, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 0.6384, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.948905109489051, | |
| "grad_norm": 0.3607243597507477, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 0.6496, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.9635036496350365, | |
| "grad_norm": 0.24368129670619965, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 0.6429, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.9781021897810219, | |
| "grad_norm": 1.090360403060913, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.7193, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.9927007299270073, | |
| "grad_norm": 0.4807930588722229, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 0.6403, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.010948905109489, | |
| "grad_norm": 1.2754051685333252, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 1.2307, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.0255474452554745, | |
| "grad_norm": 0.6335033178329468, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 0.6324, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.0401459854014599, | |
| "grad_norm": 0.8098115921020508, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 0.5842, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.0547445255474452, | |
| "grad_norm": 0.47459185123443604, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 0.6183, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.0693430656934306, | |
| "grad_norm": 0.9818742275238037, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 0.6439, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.083941605839416, | |
| "grad_norm": 0.9477861523628235, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 0.5405, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.0985401459854014, | |
| "grad_norm": 1.8204628229141235, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 0.5607, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.1131386861313868, | |
| "grad_norm": 2.8951661586761475, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 0.7007, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.1277372262773722, | |
| "grad_norm": 1.75932776927948, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 0.5977, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.1423357664233578, | |
| "grad_norm": 1.4874998331069946, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 0.8672, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.1569343065693432, | |
| "grad_norm": 1.1275254487991333, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 0.6145, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.1715328467153285, | |
| "grad_norm": 0.588616132736206, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 0.5931, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.186131386861314, | |
| "grad_norm": 0.5191588997840881, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 0.6296, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2007299270072993, | |
| "grad_norm": 0.776644229888916, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 0.5831, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.2153284671532847, | |
| "grad_norm": 0.8567450642585754, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 0.6257, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.22992700729927, | |
| "grad_norm": 0.9636512398719788, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 0.5479, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.2445255474452555, | |
| "grad_norm": 1.2442514896392822, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 0.5543, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.2591240875912408, | |
| "grad_norm": 2.7604684829711914, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.6211, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.2737226277372262, | |
| "grad_norm": 2.9896342754364014, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 0.4459, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.2883211678832116, | |
| "grad_norm": 2.8760337829589844, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 0.6636, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.3029197080291972, | |
| "grad_norm": 3.1063735485076904, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 0.5903, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.3175182481751824, | |
| "grad_norm": 2.3436295986175537, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.5067, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.332116788321168, | |
| "grad_norm": 1.832503318786621, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 0.5966, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.3467153284671534, | |
| "grad_norm": 1.6309736967086792, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 0.4747, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.3613138686131387, | |
| "grad_norm": 1.4411300420761108, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 0.5607, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.3759124087591241, | |
| "grad_norm": 1.1576124429702759, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 0.4929, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.3905109489051095, | |
| "grad_norm": 1.752060055732727, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 0.601, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.405109489051095, | |
| "grad_norm": 1.2885116338729858, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 0.5715, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.4197080291970803, | |
| "grad_norm": 1.4122116565704346, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 0.4723, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.4343065693430657, | |
| "grad_norm": 2.1593947410583496, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 0.5248, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.448905109489051, | |
| "grad_norm": 1.843381643295288, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 0.5961, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.4635036496350364, | |
| "grad_norm": 1.5842969417572021, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 0.5352, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.4635036496350364, | |
| "eval_loss": 0.44356289505958557, | |
| "eval_runtime": 8.3718, | |
| "eval_samples_per_second": 13.737, | |
| "eval_steps_per_second": 3.464, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.4781021897810218, | |
| "grad_norm": 2.1774697303771973, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 0.692, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.4927007299270074, | |
| "grad_norm": 1.3769900798797607, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 0.4823, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.5072992700729926, | |
| "grad_norm": 1.7497941255569458, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 0.5103, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.5218978102189782, | |
| "grad_norm": 1.2811167240142822, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 0.4288, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.5364963503649633, | |
| "grad_norm": 1.523463249206543, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4009, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.551094890510949, | |
| "grad_norm": 1.529181957244873, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 0.4263, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.5656934306569343, | |
| "grad_norm": 1.35500967502594, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 0.4591, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.5802919708029197, | |
| "grad_norm": 3.1416311264038086, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 0.3924, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.594890510948905, | |
| "grad_norm": 2.108579635620117, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 0.4642, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.6094890510948905, | |
| "grad_norm": 1.594874620437622, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 0.3755, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.6240875912408759, | |
| "grad_norm": 1.2215774059295654, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 0.4665, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.6386861313868613, | |
| "grad_norm": 1.5184074640274048, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 0.4768, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.6532846715328469, | |
| "grad_norm": 2.3103785514831543, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 0.5754, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.667883211678832, | |
| "grad_norm": 3.64204740524292, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 0.4808, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.6824817518248176, | |
| "grad_norm": 2.8122432231903076, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 0.5637, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.6970802919708028, | |
| "grad_norm": 2.8947243690490723, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 0.4729, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.7116788321167884, | |
| "grad_norm": 2.1363415718078613, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 0.4652, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.7262773722627736, | |
| "grad_norm": 2.799952745437622, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 0.5897, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.7408759124087592, | |
| "grad_norm": 1.240317940711975, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 0.3954, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.7554744525547445, | |
| "grad_norm": 1.7036396265029907, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 0.472, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.77007299270073, | |
| "grad_norm": 2.329824924468994, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 0.4889, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.7846715328467153, | |
| "grad_norm": 1.4788365364074707, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 0.4583, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.7992700729927007, | |
| "grad_norm": 1.58773672580719, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 0.4166, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.813868613138686, | |
| "grad_norm": 1.4104286432266235, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.3115, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.8284671532846715, | |
| "grad_norm": 1.5017321109771729, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 0.3503, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.843065693430657, | |
| "grad_norm": 1.4300916194915771, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 0.4482, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.8576642335766422, | |
| "grad_norm": 1.2180018424987793, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 0.3264, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.8722627737226278, | |
| "grad_norm": 1.6425174474716187, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 0.3637, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.886861313868613, | |
| "grad_norm": 1.9801945686340332, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 0.2746, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.9014598540145986, | |
| "grad_norm": 2.004474401473999, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 0.516, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.916058394160584, | |
| "grad_norm": 2.710179090499878, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 0.3921, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.9306569343065694, | |
| "grad_norm": 2.900529384613037, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 0.2835, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.9452554744525548, | |
| "grad_norm": 2.484365224838257, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 0.6245, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.9598540145985401, | |
| "grad_norm": 3.357897996902466, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 0.434, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.9744525547445255, | |
| "grad_norm": 3.3186569213867188, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 0.3835, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.989051094890511, | |
| "grad_norm": 2.4162700176239014, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 0.4092, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.0072992700729926, | |
| "grad_norm": 3.713386297225952, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 0.8514, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.021897810218978, | |
| "grad_norm": 2.41644549369812, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 0.2551, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.0364963503649633, | |
| "grad_norm": 2.5493385791778564, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 0.3194, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.051094890510949, | |
| "grad_norm": 3.1456234455108643, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 0.2787, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.065693430656934, | |
| "grad_norm": 1.7035255432128906, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 0.1878, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.0802919708029197, | |
| "grad_norm": 2.2034146785736084, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 0.3997, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.094890510948905, | |
| "grad_norm": 1.4337562322616577, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.2802, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.1094890510948905, | |
| "grad_norm": 1.3683956861495972, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 0.2618, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.124087591240876, | |
| "grad_norm": 1.4168925285339355, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 0.3799, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.1386861313868613, | |
| "grad_norm": 2.6454622745513916, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 0.4446, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.153284671532847, | |
| "grad_norm": 1.4156440496444702, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 0.335, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.167883211678832, | |
| "grad_norm": 1.826469898223877, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 0.3023, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.1824817518248176, | |
| "grad_norm": 1.1584259271621704, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 0.4068, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.197080291970803, | |
| "grad_norm": 1.4826784133911133, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 0.3714, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.197080291970803, | |
| "eval_loss": 0.34630635380744934, | |
| "eval_runtime": 8.3633, | |
| "eval_samples_per_second": 13.751, | |
| "eval_steps_per_second": 3.468, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.018651406336e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |