{ "best_metric": null, "best_model_checkpoint": null, "epoch": 10.0, "global_step": 162110, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03, "learning_rate": 4.9845783727098886e-05, "loss": 3.5716, "step": 500 }, { "epoch": 0.06, "learning_rate": 4.969156745419777e-05, "loss": 3.488, "step": 1000 }, { "epoch": 0.09, "learning_rate": 4.953735118129665e-05, "loss": 3.4241, "step": 1500 }, { "epoch": 0.12, "learning_rate": 4.9383134908395535e-05, "loss": 3.4241, "step": 2000 }, { "epoch": 0.15, "learning_rate": 4.922891863549442e-05, "loss": 3.4084, "step": 2500 }, { "epoch": 0.19, "learning_rate": 4.907470236259331e-05, "loss": 3.4088, "step": 3000 }, { "epoch": 0.22, "learning_rate": 4.8920486089692185e-05, "loss": 3.3556, "step": 3500 }, { "epoch": 0.25, "learning_rate": 4.876626981679107e-05, "loss": 3.3617, "step": 4000 }, { "epoch": 0.28, "learning_rate": 4.861205354388995e-05, "loss": 3.3387, "step": 4500 }, { "epoch": 0.31, "learning_rate": 4.8457837270988835e-05, "loss": 3.3603, "step": 5000 }, { "epoch": 0.34, "learning_rate": 4.8303620998087725e-05, "loss": 3.3323, "step": 5500 }, { "epoch": 0.37, "learning_rate": 4.81494047251866e-05, "loss": 3.3145, "step": 6000 }, { "epoch": 0.4, "learning_rate": 4.799518845228549e-05, "loss": 3.3182, "step": 6500 }, { "epoch": 0.43, "learning_rate": 4.784097217938437e-05, "loss": 3.3018, "step": 7000 }, { "epoch": 0.46, "learning_rate": 4.768675590648325e-05, "loss": 3.3106, "step": 7500 }, { "epoch": 0.49, "learning_rate": 4.753253963358214e-05, "loss": 3.3128, "step": 8000 }, { "epoch": 0.52, "learning_rate": 4.737832336068102e-05, "loss": 3.2902, "step": 8500 }, { "epoch": 0.56, "learning_rate": 4.722410708777991e-05, "loss": 3.2901, "step": 9000 }, { "epoch": 0.59, "learning_rate": 4.7069890814878784e-05, "loss": 3.2667, "step": 9500 }, { "epoch": 0.62, "learning_rate": 4.6915674541977675e-05, "loss": 3.252, "step": 10000 }, { "epoch": 0.65, "learning_rate": 4.676145826907656e-05, "loss": 3.2849, "step": 10500 }, { "epoch": 0.68, "learning_rate": 4.660724199617544e-05, "loss": 3.2668, "step": 11000 }, { "epoch": 0.71, "learning_rate": 4.6453025723274324e-05, "loss": 3.2758, "step": 11500 }, { "epoch": 0.74, "learning_rate": 4.62988094503732e-05, "loss": 3.2699, "step": 12000 }, { "epoch": 0.77, "learning_rate": 4.614459317747209e-05, "loss": 3.2566, "step": 12500 }, { "epoch": 0.8, "learning_rate": 4.599037690457097e-05, "loss": 3.2478, "step": 13000 }, { "epoch": 0.83, "learning_rate": 4.583616063166986e-05, "loss": 3.2408, "step": 13500 }, { "epoch": 0.86, "learning_rate": 4.568194435876874e-05, "loss": 3.2459, "step": 14000 }, { "epoch": 0.89, "learning_rate": 4.5527728085867624e-05, "loss": 3.2581, "step": 14500 }, { "epoch": 0.93, "learning_rate": 4.537351181296651e-05, "loss": 3.2349, "step": 15000 }, { "epoch": 0.96, "learning_rate": 4.521929554006539e-05, "loss": 3.2619, "step": 15500 }, { "epoch": 0.99, "learning_rate": 4.5065079267164274e-05, "loss": 3.2307, "step": 16000 }, { "epoch": 1.02, "learning_rate": 4.491086299426316e-05, "loss": 3.0671, "step": 16500 }, { "epoch": 1.05, "learning_rate": 4.475664672136204e-05, "loss": 2.9847, "step": 17000 }, { "epoch": 1.08, "learning_rate": 4.4602430448460924e-05, "loss": 2.969, "step": 17500 }, { "epoch": 1.11, "learning_rate": 4.444821417555981e-05, "loss": 2.9732, "step": 18000 }, { "epoch": 1.14, "learning_rate": 4.429399790265869e-05, "loss": 2.9817, "step": 18500 }, { "epoch": 1.17, "learning_rate": 4.413978162975757e-05, "loss": 3.0009, "step": 19000 }, { "epoch": 1.2, "learning_rate": 4.398556535685646e-05, "loss": 2.9889, "step": 19500 }, { "epoch": 1.23, "learning_rate": 4.383134908395534e-05, "loss": 2.9974, "step": 20000 }, { "epoch": 1.26, "learning_rate": 4.367713281105422e-05, "loss": 2.997, "step": 20500 }, { "epoch": 1.3, "learning_rate": 4.3522916538153106e-05, "loss": 2.9957, "step": 21000 }, { "epoch": 1.33, "learning_rate": 4.3368700265251996e-05, "loss": 3.0038, "step": 21500 }, { "epoch": 1.36, "learning_rate": 4.321448399235087e-05, "loss": 3.021, "step": 22000 }, { "epoch": 1.39, "learning_rate": 4.3060267719449756e-05, "loss": 3.0072, "step": 22500 }, { "epoch": 1.42, "learning_rate": 4.290605144654864e-05, "loss": 3.0103, "step": 23000 }, { "epoch": 1.45, "learning_rate": 4.275183517364752e-05, "loss": 3.0178, "step": 23500 }, { "epoch": 1.48, "learning_rate": 4.259761890074641e-05, "loss": 2.9975, "step": 24000 }, { "epoch": 1.51, "learning_rate": 4.244340262784529e-05, "loss": 2.9901, "step": 24500 }, { "epoch": 1.54, "learning_rate": 4.228918635494418e-05, "loss": 3.013, "step": 25000 }, { "epoch": 1.57, "learning_rate": 4.2134970082043056e-05, "loss": 3.0131, "step": 25500 }, { "epoch": 1.6, "learning_rate": 4.1980753809141946e-05, "loss": 3.0112, "step": 26000 }, { "epoch": 1.63, "learning_rate": 4.182653753624083e-05, "loss": 3.019, "step": 26500 }, { "epoch": 1.67, "learning_rate": 4.1672321263339706e-05, "loss": 3.0125, "step": 27000 }, { "epoch": 1.7, "learning_rate": 4.1518104990438596e-05, "loss": 3.0208, "step": 27500 }, { "epoch": 1.73, "learning_rate": 4.136388871753747e-05, "loss": 3.0321, "step": 28000 }, { "epoch": 1.76, "learning_rate": 4.120967244463636e-05, "loss": 3.0425, "step": 28500 }, { "epoch": 1.79, "learning_rate": 4.1055456171735245e-05, "loss": 3.0076, "step": 29000 }, { "epoch": 1.82, "learning_rate": 4.090123989883413e-05, "loss": 3.0243, "step": 29500 }, { "epoch": 1.85, "learning_rate": 4.074702362593301e-05, "loss": 3.038, "step": 30000 }, { "epoch": 1.88, "learning_rate": 4.0592807353031895e-05, "loss": 3.0264, "step": 30500 }, { "epoch": 1.91, "learning_rate": 4.043859108013078e-05, "loss": 3.0321, "step": 31000 }, { "epoch": 1.94, "learning_rate": 4.028437480722966e-05, "loss": 3.0399, "step": 31500 }, { "epoch": 1.97, "learning_rate": 4.0130158534328545e-05, "loss": 3.0232, "step": 32000 }, { "epoch": 2.0, "learning_rate": 3.997594226142743e-05, "loss": 2.9806, "step": 32500 }, { "epoch": 2.04, "learning_rate": 3.982172598852631e-05, "loss": 2.761, "step": 33000 }, { "epoch": 2.07, "learning_rate": 3.9667509715625195e-05, "loss": 2.7645, "step": 33500 }, { "epoch": 2.1, "learning_rate": 3.951329344272408e-05, "loss": 2.7481, "step": 34000 }, { "epoch": 2.13, "learning_rate": 3.935907716982296e-05, "loss": 2.7584, "step": 34500 }, { "epoch": 2.16, "learning_rate": 3.9204860896921845e-05, "loss": 2.7846, "step": 35000 }, { "epoch": 2.19, "learning_rate": 3.905064462402073e-05, "loss": 2.7609, "step": 35500 }, { "epoch": 2.22, "learning_rate": 3.889642835111961e-05, "loss": 2.7771, "step": 36000 }, { "epoch": 2.25, "learning_rate": 3.8742212078218494e-05, "loss": 2.8008, "step": 36500 }, { "epoch": 2.28, "learning_rate": 3.858799580531738e-05, "loss": 2.8029, "step": 37000 }, { "epoch": 2.31, "learning_rate": 3.843377953241626e-05, "loss": 2.7776, "step": 37500 }, { "epoch": 2.34, "learning_rate": 3.8279563259515144e-05, "loss": 2.7912, "step": 38000 }, { "epoch": 2.37, "learning_rate": 3.812534698661403e-05, "loss": 2.7857, "step": 38500 }, { "epoch": 2.41, "learning_rate": 3.797113071371291e-05, "loss": 2.8002, "step": 39000 }, { "epoch": 2.44, "learning_rate": 3.7816914440811794e-05, "loss": 2.8027, "step": 39500 }, { "epoch": 2.47, "learning_rate": 3.7662698167910684e-05, "loss": 2.8114, "step": 40000 }, { "epoch": 2.5, "learning_rate": 3.750848189500956e-05, "loss": 2.8027, "step": 40500 }, { "epoch": 2.53, "learning_rate": 3.735426562210845e-05, "loss": 2.8086, "step": 41000 }, { "epoch": 2.56, "learning_rate": 3.720004934920733e-05, "loss": 2.8143, "step": 41500 }, { "epoch": 2.59, "learning_rate": 3.704583307630621e-05, "loss": 2.8259, "step": 42000 }, { "epoch": 2.62, "learning_rate": 3.68916168034051e-05, "loss": 2.8417, "step": 42500 }, { "epoch": 2.65, "learning_rate": 3.673740053050398e-05, "loss": 2.8076, "step": 43000 }, { "epoch": 2.68, "learning_rate": 3.658318425760287e-05, "loss": 2.8147, "step": 43500 }, { "epoch": 2.71, "learning_rate": 3.6428967984701744e-05, "loss": 2.811, "step": 44000 }, { "epoch": 2.75, "learning_rate": 3.6274751711800634e-05, "loss": 2.8112, "step": 44500 }, { "epoch": 2.78, "learning_rate": 3.612053543889952e-05, "loss": 2.8338, "step": 45000 }, { "epoch": 2.81, "learning_rate": 3.596631916599839e-05, "loss": 2.8319, "step": 45500 }, { "epoch": 2.84, "learning_rate": 3.581210289309728e-05, "loss": 2.8619, "step": 46000 }, { "epoch": 2.87, "learning_rate": 3.565788662019616e-05, "loss": 2.837, "step": 46500 }, { "epoch": 2.9, "learning_rate": 3.550367034729505e-05, "loss": 2.8362, "step": 47000 }, { "epoch": 2.93, "learning_rate": 3.534945407439393e-05, "loss": 2.8691, "step": 47500 }, { "epoch": 2.96, "learning_rate": 3.5195237801492816e-05, "loss": 2.8708, "step": 48000 }, { "epoch": 2.99, "learning_rate": 3.50410215285917e-05, "loss": 2.8552, "step": 48500 }, { "epoch": 3.02, "learning_rate": 3.488680525569058e-05, "loss": 2.6518, "step": 49000 }, { "epoch": 3.05, "learning_rate": 3.4732588982789466e-05, "loss": 2.5816, "step": 49500 }, { "epoch": 3.08, "learning_rate": 3.457837270988835e-05, "loss": 2.5747, "step": 50000 }, { "epoch": 3.12, "learning_rate": 3.442415643698723e-05, "loss": 2.5805, "step": 50500 }, { "epoch": 3.15, "learning_rate": 3.4269940164086116e-05, "loss": 2.5997, "step": 51000 }, { "epoch": 3.18, "learning_rate": 3.4115723891185e-05, "loss": 2.5675, "step": 51500 }, { "epoch": 3.21, "learning_rate": 3.396150761828388e-05, "loss": 2.5793, "step": 52000 }, { "epoch": 3.24, "learning_rate": 3.3807291345382766e-05, "loss": 2.6122, "step": 52500 }, { "epoch": 3.27, "learning_rate": 3.365307507248165e-05, "loss": 2.6107, "step": 53000 }, { "epoch": 3.3, "learning_rate": 3.349885879958053e-05, "loss": 2.6113, "step": 53500 }, { "epoch": 3.33, "learning_rate": 3.3344642526679416e-05, "loss": 2.6212, "step": 54000 }, { "epoch": 3.36, "learning_rate": 3.31904262537783e-05, "loss": 2.6346, "step": 54500 }, { "epoch": 3.39, "learning_rate": 3.303620998087719e-05, "loss": 2.6517, "step": 55000 }, { "epoch": 3.42, "learning_rate": 3.2881993707976065e-05, "loss": 2.6164, "step": 55500 }, { "epoch": 3.45, "learning_rate": 3.2727777435074955e-05, "loss": 2.6346, "step": 56000 }, { "epoch": 3.49, "learning_rate": 3.257356116217383e-05, "loss": 2.6645, "step": 56500 }, { "epoch": 3.52, "learning_rate": 3.2419344889272715e-05, "loss": 2.658, "step": 57000 }, { "epoch": 3.55, "learning_rate": 3.2265128616371605e-05, "loss": 2.649, "step": 57500 }, { "epoch": 3.58, "learning_rate": 3.211091234347048e-05, "loss": 2.654, "step": 58000 }, { "epoch": 3.61, "learning_rate": 3.195669607056937e-05, "loss": 2.6365, "step": 58500 }, { "epoch": 3.64, "learning_rate": 3.180247979766825e-05, "loss": 2.6294, "step": 59000 }, { "epoch": 3.67, "learning_rate": 3.164826352476714e-05, "loss": 2.6531, "step": 59500 }, { "epoch": 3.7, "learning_rate": 3.1494047251866015e-05, "loss": 2.6313, "step": 60000 }, { "epoch": 3.73, "learning_rate": 3.13398309789649e-05, "loss": 2.668, "step": 60500 }, { "epoch": 3.76, "learning_rate": 3.118561470606379e-05, "loss": 2.6504, "step": 61000 }, { "epoch": 3.79, "learning_rate": 3.1031398433162665e-05, "loss": 2.6747, "step": 61500 }, { "epoch": 3.82, "learning_rate": 3.0877182160261555e-05, "loss": 2.6477, "step": 62000 }, { "epoch": 3.86, "learning_rate": 3.072296588736043e-05, "loss": 2.6616, "step": 62500 }, { "epoch": 3.89, "learning_rate": 3.056874961445932e-05, "loss": 2.6733, "step": 63000 }, { "epoch": 3.92, "learning_rate": 3.0414533341558205e-05, "loss": 2.6793, "step": 63500 }, { "epoch": 3.95, "learning_rate": 3.0260317068657084e-05, "loss": 2.6563, "step": 64000 }, { "epoch": 3.98, "learning_rate": 3.010610079575597e-05, "loss": 2.6447, "step": 64500 }, { "epoch": 4.01, "learning_rate": 2.995188452285485e-05, "loss": 2.6038, "step": 65000 }, { "epoch": 4.04, "learning_rate": 2.9797668249953738e-05, "loss": 2.4287, "step": 65500 }, { "epoch": 4.07, "learning_rate": 2.964345197705262e-05, "loss": 2.441, "step": 66000 }, { "epoch": 4.1, "learning_rate": 2.94892357041515e-05, "loss": 2.4128, "step": 66500 }, { "epoch": 4.13, "learning_rate": 2.9335019431250387e-05, "loss": 2.4571, "step": 67000 }, { "epoch": 4.16, "learning_rate": 2.9180803158349267e-05, "loss": 2.4115, "step": 67500 }, { "epoch": 4.19, "learning_rate": 2.9026586885448154e-05, "loss": 2.4405, "step": 68000 }, { "epoch": 4.23, "learning_rate": 2.887237061254704e-05, "loss": 2.4474, "step": 68500 }, { "epoch": 4.26, "learning_rate": 2.871815433964592e-05, "loss": 2.4561, "step": 69000 }, { "epoch": 4.29, "learning_rate": 2.8563938066744804e-05, "loss": 2.4527, "step": 69500 }, { "epoch": 4.32, "learning_rate": 2.8409721793843687e-05, "loss": 2.4407, "step": 70000 }, { "epoch": 4.35, "learning_rate": 2.825550552094257e-05, "loss": 2.4659, "step": 70500 }, { "epoch": 4.38, "learning_rate": 2.8101289248041457e-05, "loss": 2.4568, "step": 71000 }, { "epoch": 4.41, "learning_rate": 2.7947072975140337e-05, "loss": 2.4747, "step": 71500 }, { "epoch": 4.44, "learning_rate": 2.7792856702239223e-05, "loss": 2.475, "step": 72000 }, { "epoch": 4.47, "learning_rate": 2.7638640429338103e-05, "loss": 2.4684, "step": 72500 }, { "epoch": 4.5, "learning_rate": 2.748442415643699e-05, "loss": 2.5052, "step": 73000 }, { "epoch": 4.53, "learning_rate": 2.7330207883535873e-05, "loss": 2.4652, "step": 73500 }, { "epoch": 4.56, "learning_rate": 2.7175991610634753e-05, "loss": 2.4664, "step": 74000 }, { "epoch": 4.6, "learning_rate": 2.702177533773364e-05, "loss": 2.4976, "step": 74500 }, { "epoch": 4.63, "learning_rate": 2.686755906483252e-05, "loss": 2.4907, "step": 75000 }, { "epoch": 4.66, "learning_rate": 2.6713342791931406e-05, "loss": 2.5042, "step": 75500 }, { "epoch": 4.69, "learning_rate": 2.6559126519030293e-05, "loss": 2.4854, "step": 76000 }, { "epoch": 4.72, "learning_rate": 2.6404910246129173e-05, "loss": 2.4824, "step": 76500 }, { "epoch": 4.75, "learning_rate": 2.6250693973228056e-05, "loss": 2.5114, "step": 77000 }, { "epoch": 4.78, "learning_rate": 2.609647770032694e-05, "loss": 2.4997, "step": 77500 }, { "epoch": 4.81, "learning_rate": 2.5942261427425823e-05, "loss": 2.49, "step": 78000 }, { "epoch": 4.84, "learning_rate": 2.578804515452471e-05, "loss": 2.4856, "step": 78500 }, { "epoch": 4.87, "learning_rate": 2.563382888162359e-05, "loss": 2.5155, "step": 79000 }, { "epoch": 4.9, "learning_rate": 2.5479612608722476e-05, "loss": 2.4911, "step": 79500 }, { "epoch": 4.93, "learning_rate": 2.5325396335821356e-05, "loss": 2.498, "step": 80000 }, { "epoch": 4.97, "learning_rate": 2.5171180062920242e-05, "loss": 2.5189, "step": 80500 }, { "epoch": 5.0, "learning_rate": 2.5016963790019126e-05, "loss": 2.5012, "step": 81000 }, { "epoch": 5.03, "learning_rate": 2.4862747517118006e-05, "loss": 2.2808, "step": 81500 }, { "epoch": 5.06, "learning_rate": 2.470853124421689e-05, "loss": 2.2686, "step": 82000 }, { "epoch": 5.09, "learning_rate": 2.4554314971315775e-05, "loss": 2.2863, "step": 82500 }, { "epoch": 5.12, "learning_rate": 2.440009869841466e-05, "loss": 2.2829, "step": 83000 }, { "epoch": 5.15, "learning_rate": 2.4245882425513542e-05, "loss": 2.2903, "step": 83500 }, { "epoch": 5.18, "learning_rate": 2.4091666152612425e-05, "loss": 2.3022, "step": 84000 }, { "epoch": 5.21, "learning_rate": 2.393744987971131e-05, "loss": 2.3068, "step": 84500 }, { "epoch": 5.24, "learning_rate": 2.3783233606810192e-05, "loss": 2.3022, "step": 85000 }, { "epoch": 5.27, "learning_rate": 2.3629017333909075e-05, "loss": 2.3168, "step": 85500 }, { "epoch": 5.31, "learning_rate": 2.347480106100796e-05, "loss": 2.3058, "step": 86000 }, { "epoch": 5.34, "learning_rate": 2.332058478810684e-05, "loss": 2.3049, "step": 86500 }, { "epoch": 5.37, "learning_rate": 2.3166368515205725e-05, "loss": 2.2949, "step": 87000 }, { "epoch": 5.4, "learning_rate": 2.301215224230461e-05, "loss": 2.3447, "step": 87500 }, { "epoch": 5.43, "learning_rate": 2.2857935969403495e-05, "loss": 2.3235, "step": 88000 }, { "epoch": 5.46, "learning_rate": 2.2703719696502375e-05, "loss": 2.3185, "step": 88500 }, { "epoch": 5.49, "learning_rate": 2.2549503423601258e-05, "loss": 2.3146, "step": 89000 }, { "epoch": 5.52, "learning_rate": 2.239528715070014e-05, "loss": 2.3479, "step": 89500 }, { "epoch": 5.55, "learning_rate": 2.2241070877799024e-05, "loss": 2.3553, "step": 90000 }, { "epoch": 5.58, "learning_rate": 2.208685460489791e-05, "loss": 2.3444, "step": 90500 }, { "epoch": 5.61, "learning_rate": 2.1932638331996794e-05, "loss": 2.3379, "step": 91000 }, { "epoch": 5.64, "learning_rate": 2.1778422059095678e-05, "loss": 2.3675, "step": 91500 }, { "epoch": 5.68, "learning_rate": 2.162420578619456e-05, "loss": 2.346, "step": 92000 }, { "epoch": 5.71, "learning_rate": 2.1469989513293444e-05, "loss": 2.3542, "step": 92500 }, { "epoch": 5.74, "learning_rate": 2.1315773240392327e-05, "loss": 2.35, "step": 93000 }, { "epoch": 5.77, "learning_rate": 2.116155696749121e-05, "loss": 2.3474, "step": 93500 }, { "epoch": 5.8, "learning_rate": 2.1007340694590094e-05, "loss": 2.352, "step": 94000 }, { "epoch": 5.83, "learning_rate": 2.0853124421688977e-05, "loss": 2.3325, "step": 94500 }, { "epoch": 5.86, "learning_rate": 2.069890814878786e-05, "loss": 2.3424, "step": 95000 }, { "epoch": 5.89, "learning_rate": 2.0544691875886747e-05, "loss": 2.3506, "step": 95500 }, { "epoch": 5.92, "learning_rate": 2.0390475602985627e-05, "loss": 2.3756, "step": 96000 }, { "epoch": 5.95, "learning_rate": 2.023625933008451e-05, "loss": 2.3787, "step": 96500 }, { "epoch": 5.98, "learning_rate": 2.0082043057183394e-05, "loss": 2.3619, "step": 97000 }, { "epoch": 6.01, "learning_rate": 1.9927826784282277e-05, "loss": 2.2638, "step": 97500 }, { "epoch": 6.05, "learning_rate": 1.9773610511381164e-05, "loss": 2.1504, "step": 98000 }, { "epoch": 6.08, "learning_rate": 1.9619394238480047e-05, "loss": 2.141, "step": 98500 }, { "epoch": 6.11, "learning_rate": 1.946517796557893e-05, "loss": 2.1287, "step": 99000 }, { "epoch": 6.14, "learning_rate": 1.9310961692677813e-05, "loss": 2.1477, "step": 99500 }, { "epoch": 6.17, "learning_rate": 1.9156745419776693e-05, "loss": 2.1732, "step": 100000 }, { "epoch": 6.2, "learning_rate": 1.9002529146875576e-05, "loss": 2.1753, "step": 100500 }, { "epoch": 6.23, "learning_rate": 1.8848312873974463e-05, "loss": 2.1794, "step": 101000 }, { "epoch": 6.26, "learning_rate": 1.8694096601073346e-05, "loss": 2.1646, "step": 101500 }, { "epoch": 6.29, "learning_rate": 1.853988032817223e-05, "loss": 2.1717, "step": 102000 }, { "epoch": 6.32, "learning_rate": 1.8385664055271113e-05, "loss": 2.1759, "step": 102500 }, { "epoch": 6.35, "learning_rate": 1.8231447782369996e-05, "loss": 2.1716, "step": 103000 }, { "epoch": 6.38, "learning_rate": 1.807723150946888e-05, "loss": 2.1785, "step": 103500 }, { "epoch": 6.42, "learning_rate": 1.7923015236567763e-05, "loss": 2.1945, "step": 104000 }, { "epoch": 6.45, "learning_rate": 1.7768798963666646e-05, "loss": 2.2027, "step": 104500 }, { "epoch": 6.48, "learning_rate": 1.761458269076553e-05, "loss": 2.1936, "step": 105000 }, { "epoch": 6.51, "learning_rate": 1.7460366417864413e-05, "loss": 2.1835, "step": 105500 }, { "epoch": 6.54, "learning_rate": 1.73061501449633e-05, "loss": 2.2049, "step": 106000 }, { "epoch": 6.57, "learning_rate": 1.7151933872062182e-05, "loss": 2.2161, "step": 106500 }, { "epoch": 6.6, "learning_rate": 1.6997717599161066e-05, "loss": 2.2136, "step": 107000 }, { "epoch": 6.63, "learning_rate": 1.6843501326259946e-05, "loss": 2.1981, "step": 107500 }, { "epoch": 6.66, "learning_rate": 1.668928505335883e-05, "loss": 2.2208, "step": 108000 }, { "epoch": 6.69, "learning_rate": 1.6535068780457716e-05, "loss": 2.2148, "step": 108500 }, { "epoch": 6.72, "learning_rate": 1.63808525075566e-05, "loss": 2.2163, "step": 109000 }, { "epoch": 6.75, "learning_rate": 1.6226636234655482e-05, "loss": 2.202, "step": 109500 }, { "epoch": 6.79, "learning_rate": 1.6072419961754365e-05, "loss": 2.2127, "step": 110000 }, { "epoch": 6.82, "learning_rate": 1.591820368885325e-05, "loss": 2.2334, "step": 110500 }, { "epoch": 6.85, "learning_rate": 1.5763987415952132e-05, "loss": 2.2174, "step": 111000 }, { "epoch": 6.88, "learning_rate": 1.5609771143051015e-05, "loss": 2.2229, "step": 111500 }, { "epoch": 6.91, "learning_rate": 1.54555548701499e-05, "loss": 2.2225, "step": 112000 }, { "epoch": 6.94, "learning_rate": 1.5301338597248782e-05, "loss": 2.2266, "step": 112500 }, { "epoch": 6.97, "learning_rate": 1.5147122324347665e-05, "loss": 2.2197, "step": 113000 }, { "epoch": 7.0, "learning_rate": 1.4992906051446548e-05, "loss": 2.2349, "step": 113500 }, { "epoch": 7.03, "learning_rate": 1.4838689778545433e-05, "loss": 2.0295, "step": 114000 }, { "epoch": 7.06, "learning_rate": 1.4684473505644316e-05, "loss": 2.0302, "step": 114500 }, { "epoch": 7.09, "learning_rate": 1.45302572327432e-05, "loss": 2.0477, "step": 115000 }, { "epoch": 7.12, "learning_rate": 1.4376040959842083e-05, "loss": 2.0597, "step": 115500 }, { "epoch": 7.16, "learning_rate": 1.4221824686940966e-05, "loss": 2.046, "step": 116000 }, { "epoch": 7.19, "learning_rate": 1.4067608414039851e-05, "loss": 2.0449, "step": 116500 }, { "epoch": 7.22, "learning_rate": 1.3913392141138735e-05, "loss": 2.061, "step": 117000 }, { "epoch": 7.25, "learning_rate": 1.3759175868237618e-05, "loss": 2.0655, "step": 117500 }, { "epoch": 7.28, "learning_rate": 1.36049595953365e-05, "loss": 2.0623, "step": 118000 }, { "epoch": 7.31, "learning_rate": 1.3450743322435383e-05, "loss": 2.0526, "step": 118500 }, { "epoch": 7.34, "learning_rate": 1.329652704953427e-05, "loss": 2.06, "step": 119000 }, { "epoch": 7.37, "learning_rate": 1.314231077663315e-05, "loss": 2.0729, "step": 119500 }, { "epoch": 7.4, "learning_rate": 1.2988094503732034e-05, "loss": 2.0705, "step": 120000 }, { "epoch": 7.43, "learning_rate": 1.2833878230830917e-05, "loss": 2.0842, "step": 120500 }, { "epoch": 7.46, "learning_rate": 1.26796619579298e-05, "loss": 2.0781, "step": 121000 }, { "epoch": 7.49, "learning_rate": 1.2525445685028686e-05, "loss": 2.0875, "step": 121500 }, { "epoch": 7.53, "learning_rate": 1.2371229412127567e-05, "loss": 2.0942, "step": 122000 }, { "epoch": 7.56, "learning_rate": 1.2217013139226452e-05, "loss": 2.0806, "step": 122500 }, { "epoch": 7.59, "learning_rate": 1.2062796866325335e-05, "loss": 2.1183, "step": 123000 }, { "epoch": 7.62, "learning_rate": 1.1908580593424219e-05, "loss": 2.0795, "step": 123500 }, { "epoch": 7.65, "learning_rate": 1.1754364320523102e-05, "loss": 2.076, "step": 124000 }, { "epoch": 7.68, "learning_rate": 1.1600148047621985e-05, "loss": 2.0863, "step": 124500 }, { "epoch": 7.71, "learning_rate": 1.144593177472087e-05, "loss": 2.0943, "step": 125000 }, { "epoch": 7.74, "learning_rate": 1.1291715501819752e-05, "loss": 2.093, "step": 125500 }, { "epoch": 7.77, "learning_rate": 1.1137499228918637e-05, "loss": 2.1124, "step": 126000 }, { "epoch": 7.8, "learning_rate": 1.098328295601752e-05, "loss": 2.0669, "step": 126500 }, { "epoch": 7.83, "learning_rate": 1.0829066683116403e-05, "loss": 2.0848, "step": 127000 }, { "epoch": 7.87, "learning_rate": 1.0674850410215287e-05, "loss": 2.1032, "step": 127500 }, { "epoch": 7.9, "learning_rate": 1.052063413731417e-05, "loss": 2.0905, "step": 128000 }, { "epoch": 7.93, "learning_rate": 1.0366417864413053e-05, "loss": 2.0861, "step": 128500 }, { "epoch": 7.96, "learning_rate": 1.0212201591511936e-05, "loss": 2.1257, "step": 129000 }, { "epoch": 7.99, "learning_rate": 1.005798531861082e-05, "loss": 2.1051, "step": 129500 }, { "epoch": 8.02, "learning_rate": 9.903769045709705e-06, "loss": 2.0009, "step": 130000 }, { "epoch": 8.05, "learning_rate": 9.749552772808588e-06, "loss": 1.9365, "step": 130500 }, { "epoch": 8.08, "learning_rate": 9.59533649990747e-06, "loss": 1.9696, "step": 131000 }, { "epoch": 8.11, "learning_rate": 9.441120227006354e-06, "loss": 1.9647, "step": 131500 }, { "epoch": 8.14, "learning_rate": 9.286903954105238e-06, "loss": 1.969, "step": 132000 }, { "epoch": 8.17, "learning_rate": 9.132687681204123e-06, "loss": 1.9503, "step": 132500 }, { "epoch": 8.2, "learning_rate": 8.978471408303004e-06, "loss": 1.9559, "step": 133000 }, { "epoch": 8.24, "learning_rate": 8.824255135401887e-06, "loss": 1.9534, "step": 133500 }, { "epoch": 8.27, "learning_rate": 8.670038862500772e-06, "loss": 1.9658, "step": 134000 }, { "epoch": 8.3, "learning_rate": 8.515822589599656e-06, "loss": 1.9844, "step": 134500 }, { "epoch": 8.33, "learning_rate": 8.361606316698537e-06, "loss": 1.9706, "step": 135000 }, { "epoch": 8.36, "learning_rate": 8.207390043797422e-06, "loss": 1.9742, "step": 135500 }, { "epoch": 8.39, "learning_rate": 8.053173770896305e-06, "loss": 1.9711, "step": 136000 }, { "epoch": 8.42, "learning_rate": 7.898957497995189e-06, "loss": 1.9776, "step": 136500 }, { "epoch": 8.45, "learning_rate": 7.744741225094072e-06, "loss": 1.9803, "step": 137000 }, { "epoch": 8.48, "learning_rate": 7.590524952192955e-06, "loss": 1.994, "step": 137500 }, { "epoch": 8.51, "learning_rate": 7.436308679291839e-06, "loss": 1.9877, "step": 138000 }, { "epoch": 8.54, "learning_rate": 7.282092406390723e-06, "loss": 2.0027, "step": 138500 }, { "epoch": 8.57, "learning_rate": 7.127876133489605e-06, "loss": 1.9775, "step": 139000 }, { "epoch": 8.61, "learning_rate": 6.97365986058849e-06, "loss": 1.987, "step": 139500 }, { "epoch": 8.64, "learning_rate": 6.8194435876873724e-06, "loss": 2.0007, "step": 140000 }, { "epoch": 8.67, "learning_rate": 6.665227314786257e-06, "loss": 1.9846, "step": 140500 }, { "epoch": 8.7, "learning_rate": 6.51101104188514e-06, "loss": 1.9958, "step": 141000 }, { "epoch": 8.73, "learning_rate": 6.356794768984023e-06, "loss": 1.989, "step": 141500 }, { "epoch": 8.76, "learning_rate": 6.202578496082907e-06, "loss": 1.988, "step": 142000 }, { "epoch": 8.79, "learning_rate": 6.0483622231817905e-06, "loss": 1.9945, "step": 142500 }, { "epoch": 8.82, "learning_rate": 5.894145950280674e-06, "loss": 1.9978, "step": 143000 }, { "epoch": 8.85, "learning_rate": 5.739929677379558e-06, "loss": 1.9756, "step": 143500 }, { "epoch": 8.88, "learning_rate": 5.585713404478441e-06, "loss": 1.9904, "step": 144000 }, { "epoch": 8.91, "learning_rate": 5.431497131577324e-06, "loss": 2.0072, "step": 144500 }, { "epoch": 8.94, "learning_rate": 5.277280858676208e-06, "loss": 1.977, "step": 145000 }, { "epoch": 8.98, "learning_rate": 5.123064585775091e-06, "loss": 1.9875, "step": 145500 }, { "epoch": 9.01, "learning_rate": 4.968848312873975e-06, "loss": 1.9618, "step": 146000 }, { "epoch": 9.04, "learning_rate": 4.8146320399728575e-06, "loss": 1.8886, "step": 146500 }, { "epoch": 9.07, "learning_rate": 4.660415767071742e-06, "loss": 1.9177, "step": 147000 }, { "epoch": 9.1, "learning_rate": 4.506199494170625e-06, "loss": 1.8996, "step": 147500 }, { "epoch": 9.13, "learning_rate": 4.351983221269509e-06, "loss": 1.9055, "step": 148000 }, { "epoch": 9.16, "learning_rate": 4.197766948368392e-06, "loss": 1.9022, "step": 148500 }, { "epoch": 9.19, "learning_rate": 4.0435506754672755e-06, "loss": 1.8831, "step": 149000 }, { "epoch": 9.22, "learning_rate": 3.889334402566159e-06, "loss": 1.8946, "step": 149500 }, { "epoch": 9.25, "learning_rate": 3.7351181296650425e-06, "loss": 1.9204, "step": 150000 }, { "epoch": 9.28, "learning_rate": 3.580901856763926e-06, "loss": 1.9099, "step": 150500 }, { "epoch": 9.31, "learning_rate": 3.426685583862809e-06, "loss": 1.9149, "step": 151000 }, { "epoch": 9.35, "learning_rate": 3.2724693109616927e-06, "loss": 1.9105, "step": 151500 }, { "epoch": 9.38, "learning_rate": 3.1182530380605764e-06, "loss": 1.8916, "step": 152000 }, { "epoch": 9.41, "learning_rate": 2.9640367651594596e-06, "loss": 1.9104, "step": 152500 }, { "epoch": 9.44, "learning_rate": 2.8098204922583433e-06, "loss": 1.9142, "step": 153000 }, { "epoch": 9.47, "learning_rate": 2.6556042193572266e-06, "loss": 1.9066, "step": 153500 }, { "epoch": 9.5, "learning_rate": 2.5013879464561103e-06, "loss": 1.9052, "step": 154000 }, { "epoch": 9.53, "learning_rate": 2.3471716735549936e-06, "loss": 1.9173, "step": 154500 }, { "epoch": 9.56, "learning_rate": 2.1929554006538773e-06, "loss": 1.8909, "step": 155000 }, { "epoch": 9.59, "learning_rate": 2.0387391277527605e-06, "loss": 1.8904, "step": 155500 }, { "epoch": 9.62, "learning_rate": 1.884522854851644e-06, "loss": 1.897, "step": 156000 }, { "epoch": 9.65, "learning_rate": 1.7303065819505277e-06, "loss": 1.9324, "step": 156500 }, { "epoch": 9.68, "learning_rate": 1.576090309049411e-06, "loss": 1.8916, "step": 157000 }, { "epoch": 9.72, "learning_rate": 1.4218740361482944e-06, "loss": 1.9087, "step": 157500 }, { "epoch": 9.75, "learning_rate": 1.267657763247178e-06, "loss": 1.9107, "step": 158000 }, { "epoch": 9.78, "learning_rate": 1.1134414903460612e-06, "loss": 1.8954, "step": 158500 }, { "epoch": 9.81, "learning_rate": 9.592252174449449e-07, "loss": 1.9147, "step": 159000 }, { "epoch": 9.84, "learning_rate": 8.050089445438284e-07, "loss": 1.8947, "step": 159500 }, { "epoch": 9.87, "learning_rate": 6.507926716427117e-07, "loss": 1.92, "step": 160000 }, { "epoch": 9.9, "learning_rate": 4.965763987415952e-07, "loss": 1.8919, "step": 160500 }, { "epoch": 9.93, "learning_rate": 3.423601258404787e-07, "loss": 1.9054, "step": 161000 }, { "epoch": 9.96, "learning_rate": 1.8814385293936217e-07, "loss": 1.9198, "step": 161500 }, { "epoch": 9.99, "learning_rate": 3.392758003824564e-08, "loss": 1.9037, "step": 162000 }, { "epoch": 10.0, "step": 162110, "total_flos": 8.002959346916352e+16, "train_loss": 2.4721177677010764, "train_runtime": 17898.7927, "train_samples_per_second": 36.226, "train_steps_per_second": 9.057 } ], "max_steps": 162110, "num_train_epochs": 10, "total_flos": 8.002959346916352e+16, "trial_name": null, "trial_params": null }