| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9781021897810218, | |
| "eval_steps": 68, | |
| "global_step": 136, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014598540145985401, | |
| "grad_norm": 6.6188509609874355, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7522, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.029197080291970802, | |
| "grad_norm": 6.574902892362282, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7555, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.043795620437956206, | |
| "grad_norm": 4.116760676302817, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7337, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.058394160583941604, | |
| "grad_norm": 6.487316576987083, | |
| "learning_rate": 4e-05, | |
| "loss": 0.752, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.072992700729927, | |
| "grad_norm": 15.034094104997072, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8254, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08759124087591241, | |
| "grad_norm": 8.121100633280987, | |
| "learning_rate": 4.9992811366328926e-05, | |
| "loss": 0.9564, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.10218978102189781, | |
| "grad_norm": 8.8258406941248, | |
| "learning_rate": 4.997124959943201e-05, | |
| "loss": 0.8104, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.11678832116788321, | |
| "grad_norm": 6.295661645639767, | |
| "learning_rate": 4.993532709928075e-05, | |
| "loss": 0.7748, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.13138686131386862, | |
| "grad_norm": 8.793469060407903, | |
| "learning_rate": 4.9885064524570665e-05, | |
| "loss": 0.7102, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.145985401459854, | |
| "grad_norm": 4.105212126064049, | |
| "learning_rate": 4.982049078084071e-05, | |
| "loss": 0.6743, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16058394160583941, | |
| "grad_norm": 2.039360680236883, | |
| "learning_rate": 4.974164300384998e-05, | |
| "loss": 0.6023, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.17518248175182483, | |
| "grad_norm": 2.0703841635505706, | |
| "learning_rate": 4.964856653822122e-05, | |
| "loss": 0.5914, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1897810218978102, | |
| "grad_norm": 2.216652524604031, | |
| "learning_rate": 4.954131491136362e-05, | |
| "loss": 0.5223, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.20437956204379562, | |
| "grad_norm": 2.068081403621057, | |
| "learning_rate": 4.9419949802689666e-05, | |
| "loss": 0.4807, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.21897810218978103, | |
| "grad_norm": 2.2137900547871485, | |
| "learning_rate": 4.92845410081439e-05, | |
| "loss": 0.481, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.23357664233576642, | |
| "grad_norm": 3.863818843522576, | |
| "learning_rate": 4.913516640006392e-05, | |
| "loss": 0.4943, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.24817518248175183, | |
| "grad_norm": 5.538121594510211, | |
| "learning_rate": 4.897191188239667e-05, | |
| "loss": 0.4939, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.26277372262773724, | |
| "grad_norm": 2.110050058368509, | |
| "learning_rate": 4.8794871341296e-05, | |
| "loss": 0.4579, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2773722627737226, | |
| "grad_norm": 1.6455776595109672, | |
| "learning_rate": 4.8604146591129485e-05, | |
| "loss": 0.4315, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.291970802919708, | |
| "grad_norm": 1.6699338063986153, | |
| "learning_rate": 4.8399847315926e-05, | |
| "loss": 0.4436, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.30656934306569344, | |
| "grad_norm": 1.6382590699916768, | |
| "learning_rate": 4.818209100629745e-05, | |
| "loss": 0.4024, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.32116788321167883, | |
| "grad_norm": 1.473696036952172, | |
| "learning_rate": 4.795100289187099e-05, | |
| "loss": 0.407, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3357664233576642, | |
| "grad_norm": 1.5431679072262796, | |
| "learning_rate": 4.7706715869270635e-05, | |
| "loss": 0.407, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.35036496350364965, | |
| "grad_norm": 1.5667303006045037, | |
| "learning_rate": 4.74493704256897e-05, | |
| "loss": 0.3976, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.36496350364963503, | |
| "grad_norm": 1.4086778511120994, | |
| "learning_rate": 4.717911455809782e-05, | |
| "loss": 0.3932, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3795620437956204, | |
| "grad_norm": 1.255371117436458, | |
| "learning_rate": 4.6896103688129385e-05, | |
| "loss": 0.3514, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.39416058394160586, | |
| "grad_norm": 1.580105274585517, | |
| "learning_rate": 4.660050057270191e-05, | |
| "loss": 0.3679, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.40875912408759124, | |
| "grad_norm": 1.5330542124088484, | |
| "learning_rate": 4.6292475210416106e-05, | |
| "loss": 0.3809, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4233576642335766, | |
| "grad_norm": 1.36659727223936, | |
| "learning_rate": 4.597220474379125e-05, | |
| "loss": 0.3396, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.43795620437956206, | |
| "grad_norm": 1.4830662677819917, | |
| "learning_rate": 4.563987335739216e-05, | |
| "loss": 0.3581, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.45255474452554745, | |
| "grad_norm": 1.579952170056348, | |
| "learning_rate": 4.5295672171906364e-05, | |
| "loss": 0.3494, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.46715328467153283, | |
| "grad_norm": 1.5223723528568243, | |
| "learning_rate": 4.49397991342324e-05, | |
| "loss": 0.3534, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.48175182481751827, | |
| "grad_norm": 1.4533190235735434, | |
| "learning_rate": 4.4572458903642354e-05, | |
| "loss": 0.35, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.49635036496350365, | |
| "grad_norm": 1.273536091739147, | |
| "learning_rate": 4.419386273408428e-05, | |
| "loss": 0.3539, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5109489051094891, | |
| "grad_norm": 1.2519284173239225, | |
| "learning_rate": 4.3804228352691935e-05, | |
| "loss": 0.3434, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5255474452554745, | |
| "grad_norm": 1.2947790542911022, | |
| "learning_rate": 4.3403779834572004e-05, | |
| "loss": 0.3446, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5401459854014599, | |
| "grad_norm": 1.4271186712542432, | |
| "learning_rate": 4.2992747473940556e-05, | |
| "loss": 0.3485, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.5547445255474452, | |
| "grad_norm": 1.1763950013950975, | |
| "learning_rate": 4.2571367651683e-05, | |
| "loss": 0.3117, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5693430656934306, | |
| "grad_norm": 1.1117128867059132, | |
| "learning_rate": 4.213988269941362e-05, | |
| "loss": 0.2661, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.583941605839416, | |
| "grad_norm": 1.2354957130115665, | |
| "learning_rate": 4.169854076011292e-05, | |
| "loss": 0.3013, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5985401459854015, | |
| "grad_norm": 1.163629462061701, | |
| "learning_rate": 4.124759564542295e-05, | |
| "loss": 0.293, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6131386861313869, | |
| "grad_norm": 1.1963752057859445, | |
| "learning_rate": 4.078730668968253e-05, | |
| "loss": 0.2704, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6277372262773723, | |
| "grad_norm": 1.1433957305246072, | |
| "learning_rate": 4.031793860078649e-05, | |
| "loss": 0.2798, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6423357664233577, | |
| "grad_norm": 1.0742244561363925, | |
| "learning_rate": 3.9839761307954675e-05, | |
| "loss": 0.2668, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.656934306569343, | |
| "grad_norm": 1.139472353645151, | |
| "learning_rate": 3.935304980649813e-05, | |
| "loss": 0.3206, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6715328467153284, | |
| "grad_norm": 1.1118011686610685, | |
| "learning_rate": 3.8858083999671855e-05, | |
| "loss": 0.28, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6861313868613139, | |
| "grad_norm": 1.105867383261182, | |
| "learning_rate": 3.835514853770505e-05, | |
| "loss": 0.2861, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7007299270072993, | |
| "grad_norm": 1.0643135219882822, | |
| "learning_rate": 3.784453265410141e-05, | |
| "loss": 0.2677, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7153284671532847, | |
| "grad_norm": 0.970941214477793, | |
| "learning_rate": 3.732652999930364e-05, | |
| "loss": 0.2419, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7299270072992701, | |
| "grad_norm": 1.2039034119373921, | |
| "learning_rate": 3.680143847181783e-05, | |
| "loss": 0.2817, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7445255474452555, | |
| "grad_norm": 1.3763613598969262, | |
| "learning_rate": 3.6269560046894766e-05, | |
| "loss": 0.2813, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.7591240875912408, | |
| "grad_norm": 1.0917986372749406, | |
| "learning_rate": 3.573120060286679e-05, | |
| "loss": 0.2434, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.7737226277372263, | |
| "grad_norm": 1.2010408408369964, | |
| "learning_rate": 3.5186669745240026e-05, | |
| "loss": 0.266, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.7883211678832117, | |
| "grad_norm": 0.9851709556799068, | |
| "learning_rate": 3.463628062864312e-05, | |
| "loss": 0.2401, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8029197080291971, | |
| "grad_norm": 1.027902148221402, | |
| "learning_rate": 3.4080349776734925e-05, | |
| "loss": 0.2396, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8175182481751825, | |
| "grad_norm": 1.038222274485028, | |
| "learning_rate": 3.351919690017473e-05, | |
| "loss": 0.2396, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8321167883211679, | |
| "grad_norm": 0.9906398418415747, | |
| "learning_rate": 3.2953144712759545e-05, | |
| "loss": 0.2435, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.8467153284671532, | |
| "grad_norm": 0.9958710262970785, | |
| "learning_rate": 3.238251874583452e-05, | |
| "loss": 0.2585, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.8613138686131386, | |
| "grad_norm": 0.9280508074490356, | |
| "learning_rate": 3.1807647161082795e-05, | |
| "loss": 0.2328, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.8759124087591241, | |
| "grad_norm": 0.9551807399682243, | |
| "learning_rate": 3.122886056180284e-05, | |
| "loss": 0.248, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8905109489051095, | |
| "grad_norm": 1.0680823587174857, | |
| "learning_rate": 3.064649180278152e-05, | |
| "loss": 0.2766, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9051094890510949, | |
| "grad_norm": 2.730287207993988, | |
| "learning_rate": 3.006087579887244e-05, | |
| "loss": 0.2274, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.9197080291970803, | |
| "grad_norm": 1.0477423658369303, | |
| "learning_rate": 2.9472349332389525e-05, | |
| "loss": 0.2163, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.9343065693430657, | |
| "grad_norm": 1.0633922453008755, | |
| "learning_rate": 2.8881250859426646e-05, | |
| "loss": 0.2552, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.948905109489051, | |
| "grad_norm": 0.940535973392031, | |
| "learning_rate": 2.8287920315214643e-05, | |
| "loss": 0.2268, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.9635036496350365, | |
| "grad_norm": 0.8335596155443756, | |
| "learning_rate": 2.7692698918627778e-05, | |
| "loss": 0.2051, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.9781021897810219, | |
| "grad_norm": 0.9709259192327347, | |
| "learning_rate": 2.7095928975951913e-05, | |
| "loss": 0.238, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.9927007299270073, | |
| "grad_norm": 0.9181878310139981, | |
| "learning_rate": 2.649795368402735e-05, | |
| "loss": 0.2607, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.9181878310139981, | |
| "learning_rate": 2.649795368402735e-05, | |
| "loss": 0.1126, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5867471694946289, | |
| "eval_runtime": 2.5685, | |
| "eval_samples_per_second": 27.643, | |
| "eval_steps_per_second": 1.168, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.0145985401459854, | |
| "grad_norm": 0.817428076729955, | |
| "learning_rate": 2.5899116932879534e-05, | |
| "loss": 0.1198, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.0291970802919708, | |
| "grad_norm": 0.7442926018853524, | |
| "learning_rate": 2.529976310795108e-05, | |
| "loss": 0.1285, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.0437956204379562, | |
| "grad_norm": 0.782523876321553, | |
| "learning_rate": 2.470023689204893e-05, | |
| "loss": 0.1272, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.0583941605839415, | |
| "grad_norm": 0.7879156799249153, | |
| "learning_rate": 2.4100883067120475e-05, | |
| "loss": 0.114, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.072992700729927, | |
| "grad_norm": 0.6738289101740561, | |
| "learning_rate": 2.3502046315972656e-05, | |
| "loss": 0.1161, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.0875912408759123, | |
| "grad_norm": 0.7921210574551151, | |
| "learning_rate": 2.290407102404809e-05, | |
| "loss": 0.1078, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.102189781021898, | |
| "grad_norm": 0.7872824090927025, | |
| "learning_rate": 2.2307301081372224e-05, | |
| "loss": 0.1311, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.1167883211678833, | |
| "grad_norm": 0.8331285832919668, | |
| "learning_rate": 2.1712079684785363e-05, | |
| "loss": 0.0932, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.1313868613138687, | |
| "grad_norm": 0.7662201225372701, | |
| "learning_rate": 2.111874914057336e-05, | |
| "loss": 0.1339, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.145985401459854, | |
| "grad_norm": 0.8748763622681786, | |
| "learning_rate": 2.0527650667610478e-05, | |
| "loss": 0.1328, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.1605839416058394, | |
| "grad_norm": 0.764695112873272, | |
| "learning_rate": 1.993912420112756e-05, | |
| "loss": 0.0968, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.1751824817518248, | |
| "grad_norm": 0.8003697459148763, | |
| "learning_rate": 1.935350819721849e-05, | |
| "loss": 0.1303, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.1897810218978102, | |
| "grad_norm": 0.8866466859845482, | |
| "learning_rate": 1.8771139438197168e-05, | |
| "loss": 0.1273, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.2043795620437956, | |
| "grad_norm": 0.7202427078538033, | |
| "learning_rate": 1.819235283891721e-05, | |
| "loss": 0.1036, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.218978102189781, | |
| "grad_norm": 0.6939994466496473, | |
| "learning_rate": 1.7617481254165487e-05, | |
| "loss": 0.1166, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.2335766423357664, | |
| "grad_norm": 0.6773010302740605, | |
| "learning_rate": 1.704685528724046e-05, | |
| "loss": 0.1239, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.2481751824817517, | |
| "grad_norm": 0.8057614185871913, | |
| "learning_rate": 1.648080309982528e-05, | |
| "loss": 0.1269, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.2627737226277373, | |
| "grad_norm": 0.738081394228655, | |
| "learning_rate": 1.591965022326507e-05, | |
| "loss": 0.1184, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.2773722627737225, | |
| "grad_norm": 0.7034144989855043, | |
| "learning_rate": 1.536371937135688e-05, | |
| "loss": 0.105, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.2919708029197081, | |
| "grad_norm": 0.809614248272801, | |
| "learning_rate": 1.4813330254759985e-05, | |
| "loss": 0.1299, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.3065693430656935, | |
| "grad_norm": 0.7359581131995093, | |
| "learning_rate": 1.426879939713322e-05, | |
| "loss": 0.0931, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.3211678832116789, | |
| "grad_norm": 0.6271750011013629, | |
| "learning_rate": 1.3730439953105243e-05, | |
| "loss": 0.1, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.3357664233576643, | |
| "grad_norm": 0.6681653257507313, | |
| "learning_rate": 1.3198561528182183e-05, | |
| "loss": 0.0882, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.3503649635036497, | |
| "grad_norm": 0.5892983169851506, | |
| "learning_rate": 1.2673470000696364e-05, | |
| "loss": 0.119, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.364963503649635, | |
| "grad_norm": 0.7781672091856281, | |
| "learning_rate": 1.2155467345898602e-05, | |
| "loss": 0.1235, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.3795620437956204, | |
| "grad_norm": 0.7457533748551523, | |
| "learning_rate": 1.1644851462294957e-05, | |
| "loss": 0.1153, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.3941605839416058, | |
| "grad_norm": 0.7719451070162108, | |
| "learning_rate": 1.114191600032815e-05, | |
| "loss": 0.1108, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.4087591240875912, | |
| "grad_norm": 0.6456397546412238, | |
| "learning_rate": 1.064695019350187e-05, | |
| "loss": 0.0966, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.4233576642335766, | |
| "grad_norm": 0.7075931478021869, | |
| "learning_rate": 1.0160238692045332e-05, | |
| "loss": 0.122, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.437956204379562, | |
| "grad_norm": 0.7164358739819071, | |
| "learning_rate": 9.682061399213525e-06, | |
| "loss": 0.1073, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.4525547445255476, | |
| "grad_norm": 0.6742795850736939, | |
| "learning_rate": 9.21269331031748e-06, | |
| "loss": 0.1082, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.4671532846715327, | |
| "grad_norm": 0.664349783270638, | |
| "learning_rate": 8.752404354577052e-06, | |
| "loss": 0.1086, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.4817518248175183, | |
| "grad_norm": 0.6104029576807147, | |
| "learning_rate": 8.301459239887074e-06, | |
| "loss": 0.095, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.4963503649635037, | |
| "grad_norm": 0.9865461135511884, | |
| "learning_rate": 7.860117300586383e-06, | |
| "loss": 0.1213, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.510948905109489, | |
| "grad_norm": 0.6388824796222101, | |
| "learning_rate": 7.428632348317005e-06, | |
| "loss": 0.082, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.5255474452554745, | |
| "grad_norm": 0.5887544819341681, | |
| "learning_rate": 7.007252526059446e-06, | |
| "loss": 0.0807, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.5401459854014599, | |
| "grad_norm": 0.5287062328074978, | |
| "learning_rate": 6.596220165428002e-06, | |
| "loss": 0.0946, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.5547445255474452, | |
| "grad_norm": 0.693208580525385, | |
| "learning_rate": 6.195771647308074e-06, | |
| "loss": 0.116, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.5693430656934306, | |
| "grad_norm": 0.6615214534660775, | |
| "learning_rate": 5.806137265915732e-06, | |
| "loss": 0.1029, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.583941605839416, | |
| "grad_norm": 0.6252962266934661, | |
| "learning_rate": 5.427541096357647e-06, | |
| "loss": 0.0999, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.5985401459854014, | |
| "grad_norm": 0.6524052423711888, | |
| "learning_rate": 5.060200865767606e-06, | |
| "loss": 0.1018, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.613138686131387, | |
| "grad_norm": 0.67321527870827, | |
| "learning_rate": 4.704327828093641e-06, | |
| "loss": 0.1015, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.6277372262773722, | |
| "grad_norm": 0.6002708673392801, | |
| "learning_rate": 4.3601266426078426e-06, | |
| "loss": 0.0919, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.6423357664233578, | |
| "grad_norm": 0.6354891189727261, | |
| "learning_rate": 4.02779525620875e-06, | |
| "loss": 0.0962, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.656934306569343, | |
| "grad_norm": 0.6211165567663863, | |
| "learning_rate": 3.707524789583891e-06, | |
| "loss": 0.0838, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.6715328467153285, | |
| "grad_norm": 0.6058146661351359, | |
| "learning_rate": 3.3994994272980946e-06, | |
| "loss": 0.0965, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.686131386861314, | |
| "grad_norm": 0.6098982651990246, | |
| "learning_rate": 3.1038963118706244e-06, | |
| "loss": 0.1097, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.7007299270072993, | |
| "grad_norm": 0.6644009843756008, | |
| "learning_rate": 2.8208854419021824e-06, | |
| "loss": 0.0903, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.7153284671532847, | |
| "grad_norm": 0.5210745143934205, | |
| "learning_rate": 2.5506295743103094e-06, | |
| "loss": 0.0717, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.72992700729927, | |
| "grad_norm": 0.6089605504043503, | |
| "learning_rate": 2.2932841307293644e-06, | |
| "loss": 0.103, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.7445255474452555, | |
| "grad_norm": 0.6090432120427985, | |
| "learning_rate": 2.0489971081290195e-06, | |
| "loss": 0.0895, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.7591240875912408, | |
| "grad_norm": 0.5490697461707983, | |
| "learning_rate": 1.817908993702555e-06, | |
| "loss": 0.0843, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.7737226277372264, | |
| "grad_norm": 0.5710673269130835, | |
| "learning_rate": 1.6001526840740049e-06, | |
| "loss": 0.0906, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.7883211678832116, | |
| "grad_norm": 0.6245928434878661, | |
| "learning_rate": 1.3958534088705206e-06, | |
| "loss": 0.0776, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.8029197080291972, | |
| "grad_norm": 0.549963273284548, | |
| "learning_rate": 1.205128658704005e-06, | |
| "loss": 0.0803, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.8175182481751824, | |
| "grad_norm": 0.6057327330974007, | |
| "learning_rate": 1.028088117603332e-06, | |
| "loss": 0.0714, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.832116788321168, | |
| "grad_norm": 0.5865246963245718, | |
| "learning_rate": 8.648335999360935e-07, | |
| "loss": 0.0967, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.8467153284671531, | |
| "grad_norm": 0.5746297279399645, | |
| "learning_rate": 7.154589918561022e-07, | |
| "loss": 0.0849, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.8613138686131387, | |
| "grad_norm": 0.600256652314103, | |
| "learning_rate": 5.800501973103362e-07, | |
| "loss": 0.0916, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.8759124087591241, | |
| "grad_norm": 0.6555734310326506, | |
| "learning_rate": 4.586850886363875e-07, | |
| "loss": 0.1204, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.8905109489051095, | |
| "grad_norm": 0.6566157654207085, | |
| "learning_rate": 3.514334617787857e-07, | |
| "loss": 0.081, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.905109489051095, | |
| "grad_norm": 0.5292071962356375, | |
| "learning_rate": 2.5835699615002764e-07, | |
| "loss": 0.0727, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.9197080291970803, | |
| "grad_norm": 0.6235513391783781, | |
| "learning_rate": 1.7950921915928788e-07, | |
| "loss": 0.0967, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.9343065693430657, | |
| "grad_norm": 0.6604731273435182, | |
| "learning_rate": 1.1493547542933969e-07, | |
| "loss": 0.1037, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.948905109489051, | |
| "grad_norm": 0.6217262060263744, | |
| "learning_rate": 6.467290071925647e-08, | |
| "loss": 0.0711, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.9635036496350367, | |
| "grad_norm": 0.535719020023098, | |
| "learning_rate": 2.8750400567992274e-08, | |
| "loss": 0.0799, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.9781021897810218, | |
| "grad_norm": 0.6249296896177375, | |
| "learning_rate": 7.188633671079137e-09, | |
| "loss": 0.097, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.9781021897810218, | |
| "eval_loss": 0.6955251693725586, | |
| "eval_runtime": 1.1203, | |
| "eval_samples_per_second": 63.374, | |
| "eval_steps_per_second": 2.678, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.9781021897810218, | |
| "step": 136, | |
| "total_flos": 79316025114624.0, | |
| "train_loss": 0.2505467278773294, | |
| "train_runtime": 551.7554, | |
| "train_samples_per_second": 19.886, | |
| "train_steps_per_second": 0.246 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 136, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 1, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 79316025114624.0, | |
| "train_batch_size": 5, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |