{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "global_step": 199995, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 4.9874996874921875e-05, "loss": 3.2939, "step": 500 }, { "epoch": 0.02, "learning_rate": 4.974999374984375e-05, "loss": 3.0017, "step": 1000 }, { "epoch": 0.02, "learning_rate": 4.962499062476562e-05, "loss": 2.8608, "step": 1500 }, { "epoch": 0.03, "learning_rate": 4.949998749968749e-05, "loss": 2.7653, "step": 2000 }, { "epoch": 0.04, "learning_rate": 4.937498437460937e-05, "loss": 2.681, "step": 2500 }, { "epoch": 0.05, "learning_rate": 4.924998124953124e-05, "loss": 2.6364, "step": 3000 }, { "epoch": 0.05, "learning_rate": 4.9124978124453116e-05, "loss": 2.5911, "step": 3500 }, { "epoch": 0.06, "learning_rate": 4.899997499937499e-05, "loss": 2.5177, "step": 4000 }, { "epoch": 0.07, "learning_rate": 4.887497187429686e-05, "loss": 2.5272, "step": 4500 }, { "epoch": 0.08, "learning_rate": 4.874996874921873e-05, "loss": 2.4444, "step": 5000 }, { "epoch": 0.08, "learning_rate": 4.8624965624140605e-05, "loss": 2.4071, "step": 5500 }, { "epoch": 0.09, "learning_rate": 4.849996249906248e-05, "loss": 2.4294, "step": 6000 }, { "epoch": 0.1, "learning_rate": 4.837495937398435e-05, "loss": 2.3753, "step": 6500 }, { "epoch": 0.11, "learning_rate": 4.824995624890622e-05, "loss": 2.3464, "step": 7000 }, { "epoch": 0.11, "learning_rate": 4.81249531238281e-05, "loss": 2.3459, "step": 7500 }, { "epoch": 0.12, "learning_rate": 4.7999949998749974e-05, "loss": 2.3276, "step": 8000 }, { "epoch": 0.13, "learning_rate": 4.7874946873671846e-05, "loss": 2.2796, "step": 8500 }, { "epoch": 0.14, "learning_rate": 4.774994374859372e-05, "loss": 2.3016, "step": 9000 }, { "epoch": 0.14, "learning_rate": 4.762494062351559e-05, "loss": 2.2536, "step": 9500 }, { "epoch": 0.15, "learning_rate": 4.749993749843746e-05, "loss": 2.2602, "step": 10000 }, { "epoch": 0.16, "learning_rate": 4.7374934373359335e-05, "loss": 2.2181, "step": 10500 }, { "epoch": 0.17, "learning_rate": 4.724993124828121e-05, "loss": 2.2444, "step": 11000 }, { "epoch": 0.17, "learning_rate": 4.712492812320308e-05, "loss": 2.2094, "step": 11500 }, { "epoch": 0.18, "learning_rate": 4.699992499812495e-05, "loss": 2.1863, "step": 12000 }, { "epoch": 0.19, "learning_rate": 4.687492187304683e-05, "loss": 2.1926, "step": 12500 }, { "epoch": 0.2, "learning_rate": 4.6749918747968704e-05, "loss": 2.1556, "step": 13000 }, { "epoch": 0.2, "learning_rate": 4.6624915622890576e-05, "loss": 2.1883, "step": 13500 }, { "epoch": 0.21, "learning_rate": 4.649991249781245e-05, "loss": 2.1666, "step": 14000 }, { "epoch": 0.22, "learning_rate": 4.637490937273432e-05, "loss": 2.1424, "step": 14500 }, { "epoch": 0.23, "learning_rate": 4.624990624765619e-05, "loss": 2.1152, "step": 15000 }, { "epoch": 0.23, "learning_rate": 4.6124903122578066e-05, "loss": 2.1066, "step": 15500 }, { "epoch": 0.24, "learning_rate": 4.599989999749994e-05, "loss": 2.1317, "step": 16000 }, { "epoch": 0.25, "learning_rate": 4.587489687242181e-05, "loss": 2.1149, "step": 16500 }, { "epoch": 0.26, "learning_rate": 4.574989374734369e-05, "loss": 2.0706, "step": 17000 }, { "epoch": 0.26, "learning_rate": 4.562489062226556e-05, "loss": 2.0609, "step": 17500 }, { "epoch": 0.27, "learning_rate": 4.5499887497187434e-05, "loss": 2.0769, "step": 18000 }, { "epoch": 0.28, "learning_rate": 4.5374884372109307e-05, "loss": 2.0559, "step": 18500 }, { "epoch": 0.29, "learning_rate": 4.524988124703118e-05, "loss": 2.0441, "step": 19000 }, { "epoch": 0.29, "learning_rate": 4.512487812195305e-05, "loss": 2.0601, "step": 19500 }, { "epoch": 0.3, "learning_rate": 4.4999874996874924e-05, "loss": 2.0693, "step": 20000 }, { "epoch": 0.31, "learning_rate": 4.4874871871796796e-05, "loss": 2.0679, "step": 20500 }, { "epoch": 0.32, "learning_rate": 4.474986874671867e-05, "loss": 2.0453, "step": 21000 }, { "epoch": 0.32, "learning_rate": 4.462486562164054e-05, "loss": 2.0092, "step": 21500 }, { "epoch": 0.33, "learning_rate": 4.449986249656242e-05, "loss": 2.0059, "step": 22000 }, { "epoch": 0.34, "learning_rate": 4.437485937148429e-05, "loss": 2.0032, "step": 22500 }, { "epoch": 0.35, "learning_rate": 4.4249856246406165e-05, "loss": 2.0177, "step": 23000 }, { "epoch": 0.35, "learning_rate": 4.412485312132804e-05, "loss": 1.9968, "step": 23500 }, { "epoch": 0.36, "learning_rate": 4.399984999624991e-05, "loss": 1.9988, "step": 24000 }, { "epoch": 0.37, "learning_rate": 4.387484687117178e-05, "loss": 1.9992, "step": 24500 }, { "epoch": 0.38, "learning_rate": 4.3749843746093654e-05, "loss": 1.9994, "step": 25000 }, { "epoch": 0.38, "learning_rate": 4.3624840621015526e-05, "loss": 1.9856, "step": 25500 }, { "epoch": 0.39, "learning_rate": 4.34998374959374e-05, "loss": 1.9882, "step": 26000 }, { "epoch": 0.4, "learning_rate": 4.337483437085927e-05, "loss": 1.9656, "step": 26500 }, { "epoch": 0.41, "learning_rate": 4.324983124578115e-05, "loss": 1.9613, "step": 27000 }, { "epoch": 0.41, "learning_rate": 4.312482812070302e-05, "loss": 1.9467, "step": 27500 }, { "epoch": 0.42, "learning_rate": 4.2999824995624895e-05, "loss": 1.9539, "step": 28000 }, { "epoch": 0.43, "learning_rate": 4.287482187054677e-05, "loss": 1.9718, "step": 28500 }, { "epoch": 0.44, "learning_rate": 4.274981874546864e-05, "loss": 1.9498, "step": 29000 }, { "epoch": 0.44, "learning_rate": 4.262481562039051e-05, "loss": 1.9421, "step": 29500 }, { "epoch": 0.45, "learning_rate": 4.2499812495312384e-05, "loss": 1.9319, "step": 30000 }, { "epoch": 0.46, "learning_rate": 4.2374809370234257e-05, "loss": 1.9388, "step": 30500 }, { "epoch": 0.47, "learning_rate": 4.224980624515613e-05, "loss": 1.905, "step": 31000 }, { "epoch": 0.47, "learning_rate": 4.2124803120078e-05, "loss": 1.9323, "step": 31500 }, { "epoch": 0.48, "learning_rate": 4.199979999499988e-05, "loss": 1.9367, "step": 32000 }, { "epoch": 0.49, "learning_rate": 4.187479686992175e-05, "loss": 1.9217, "step": 32500 }, { "epoch": 0.5, "learning_rate": 4.1749793744843625e-05, "loss": 1.9409, "step": 33000 }, { "epoch": 0.5, "learning_rate": 4.16247906197655e-05, "loss": 1.8921, "step": 33500 }, { "epoch": 0.51, "learning_rate": 4.149978749468737e-05, "loss": 1.9078, "step": 34000 }, { "epoch": 0.52, "learning_rate": 4.137478436960924e-05, "loss": 1.8943, "step": 34500 }, { "epoch": 0.53, "learning_rate": 4.1249781244531115e-05, "loss": 1.8869, "step": 35000 }, { "epoch": 0.53, "learning_rate": 4.112477811945299e-05, "loss": 1.8941, "step": 35500 }, { "epoch": 0.54, "learning_rate": 4.099977499437486e-05, "loss": 1.8851, "step": 36000 }, { "epoch": 0.55, "learning_rate": 4.087477186929674e-05, "loss": 1.8956, "step": 36500 }, { "epoch": 0.56, "learning_rate": 4.074976874421861e-05, "loss": 1.8973, "step": 37000 }, { "epoch": 0.56, "learning_rate": 4.062476561914048e-05, "loss": 1.8878, "step": 37500 }, { "epoch": 0.57, "learning_rate": 4.0499762494062355e-05, "loss": 1.8487, "step": 38000 }, { "epoch": 0.58, "learning_rate": 4.037475936898423e-05, "loss": 1.8884, "step": 38500 }, { "epoch": 0.59, "learning_rate": 4.02497562439061e-05, "loss": 1.8676, "step": 39000 }, { "epoch": 0.59, "learning_rate": 4.012475311882797e-05, "loss": 1.8499, "step": 39500 }, { "epoch": 0.6, "learning_rate": 3.9999749993749845e-05, "loss": 1.8559, "step": 40000 }, { "epoch": 0.61, "learning_rate": 3.987474686867172e-05, "loss": 1.863, "step": 40500 }, { "epoch": 0.62, "learning_rate": 3.974974374359359e-05, "loss": 1.86, "step": 41000 }, { "epoch": 0.62, "learning_rate": 3.962474061851547e-05, "loss": 1.8411, "step": 41500 }, { "epoch": 0.63, "learning_rate": 3.949973749343734e-05, "loss": 1.8469, "step": 42000 }, { "epoch": 0.64, "learning_rate": 3.9374734368359213e-05, "loss": 1.8404, "step": 42500 }, { "epoch": 0.65, "learning_rate": 3.9249731243281086e-05, "loss": 1.855, "step": 43000 }, { "epoch": 0.65, "learning_rate": 3.912472811820296e-05, "loss": 1.8326, "step": 43500 }, { "epoch": 0.66, "learning_rate": 3.899972499312483e-05, "loss": 1.8078, "step": 44000 }, { "epoch": 0.67, "learning_rate": 3.88747218680467e-05, "loss": 1.8437, "step": 44500 }, { "epoch": 0.68, "learning_rate": 3.8749718742968575e-05, "loss": 1.8437, "step": 45000 }, { "epoch": 0.68, "learning_rate": 3.862471561789045e-05, "loss": 1.8297, "step": 45500 }, { "epoch": 0.69, "learning_rate": 3.849971249281232e-05, "loss": 1.8333, "step": 46000 }, { "epoch": 0.7, "learning_rate": 3.83747093677342e-05, "loss": 1.835, "step": 46500 }, { "epoch": 0.71, "learning_rate": 3.824970624265607e-05, "loss": 1.8329, "step": 47000 }, { "epoch": 0.71, "learning_rate": 3.8124703117577944e-05, "loss": 1.8214, "step": 47500 }, { "epoch": 0.72, "learning_rate": 3.7999699992499816e-05, "loss": 1.8136, "step": 48000 }, { "epoch": 0.73, "learning_rate": 3.787469686742169e-05, "loss": 1.8138, "step": 48500 }, { "epoch": 0.74, "learning_rate": 3.774969374234356e-05, "loss": 1.7895, "step": 49000 }, { "epoch": 0.74, "learning_rate": 3.762469061726543e-05, "loss": 1.8076, "step": 49500 }, { "epoch": 0.75, "learning_rate": 3.7499687492187305e-05, "loss": 1.7935, "step": 50000 }, { "epoch": 0.76, "learning_rate": 3.737468436710918e-05, "loss": 1.8139, "step": 50500 }, { "epoch": 0.77, "learning_rate": 3.724968124203106e-05, "loss": 1.7982, "step": 51000 }, { "epoch": 0.77, "learning_rate": 3.712467811695293e-05, "loss": 1.8233, "step": 51500 }, { "epoch": 0.78, "learning_rate": 3.69996749918748e-05, "loss": 1.7931, "step": 52000 }, { "epoch": 0.79, "learning_rate": 3.6874671866796674e-05, "loss": 1.7879, "step": 52500 }, { "epoch": 0.8, "learning_rate": 3.6749668741718546e-05, "loss": 1.7863, "step": 53000 }, { "epoch": 0.8, "learning_rate": 3.662466561664042e-05, "loss": 1.7788, "step": 53500 }, { "epoch": 0.81, "learning_rate": 3.649966249156229e-05, "loss": 1.7713, "step": 54000 }, { "epoch": 0.82, "learning_rate": 3.6374659366484163e-05, "loss": 1.7693, "step": 54500 }, { "epoch": 0.83, "learning_rate": 3.6249656241406036e-05, "loss": 1.7801, "step": 55000 }, { "epoch": 0.83, "learning_rate": 3.612465311632791e-05, "loss": 1.7827, "step": 55500 }, { "epoch": 0.84, "learning_rate": 3.599964999124979e-05, "loss": 1.7424, "step": 56000 }, { "epoch": 0.85, "learning_rate": 3.587464686617166e-05, "loss": 1.783, "step": 56500 }, { "epoch": 0.86, "learning_rate": 3.574964374109353e-05, "loss": 1.7444, "step": 57000 }, { "epoch": 0.86, "learning_rate": 3.5624640616015404e-05, "loss": 1.7505, "step": 57500 }, { "epoch": 0.87, "learning_rate": 3.549963749093728e-05, "loss": 1.7765, "step": 58000 }, { "epoch": 0.88, "learning_rate": 3.537463436585915e-05, "loss": 1.7915, "step": 58500 }, { "epoch": 0.89, "learning_rate": 3.524963124078102e-05, "loss": 1.7407, "step": 59000 }, { "epoch": 0.89, "learning_rate": 3.5124628115702894e-05, "loss": 1.7584, "step": 59500 }, { "epoch": 0.9, "learning_rate": 3.4999624990624766e-05, "loss": 1.756, "step": 60000 }, { "epoch": 0.91, "learning_rate": 3.487462186554664e-05, "loss": 1.7643, "step": 60500 }, { "epoch": 0.92, "learning_rate": 3.474961874046852e-05, "loss": 1.7417, "step": 61000 }, { "epoch": 0.92, "learning_rate": 3.462461561539039e-05, "loss": 1.7739, "step": 61500 }, { "epoch": 0.93, "learning_rate": 3.449961249031226e-05, "loss": 1.74, "step": 62000 }, { "epoch": 0.94, "learning_rate": 3.4374609365234135e-05, "loss": 1.7517, "step": 62500 }, { "epoch": 0.95, "learning_rate": 3.424960624015601e-05, "loss": 1.7384, "step": 63000 }, { "epoch": 0.95, "learning_rate": 3.412460311507788e-05, "loss": 1.7276, "step": 63500 }, { "epoch": 0.96, "learning_rate": 3.399959998999975e-05, "loss": 1.7359, "step": 64000 }, { "epoch": 0.97, "learning_rate": 3.3874596864921624e-05, "loss": 1.7279, "step": 64500 }, { "epoch": 0.98, "learning_rate": 3.3749593739843496e-05, "loss": 1.7402, "step": 65000 }, { "epoch": 0.98, "learning_rate": 3.3624590614765376e-05, "loss": 1.7468, "step": 65500 }, { "epoch": 0.99, "learning_rate": 3.349958748968725e-05, "loss": 1.7205, "step": 66000 }, { "epoch": 1.0, "learning_rate": 3.337458436460912e-05, "loss": 1.7191, "step": 66500 }, { "epoch": 1.01, "learning_rate": 3.324958123953099e-05, "loss": 1.6651, "step": 67000 }, { "epoch": 1.01, "learning_rate": 3.3124578114452865e-05, "loss": 1.5845, "step": 67500 }, { "epoch": 1.02, "learning_rate": 3.299957498937474e-05, "loss": 1.602, "step": 68000 }, { "epoch": 1.03, "learning_rate": 3.287457186429661e-05, "loss": 1.6055, "step": 68500 }, { "epoch": 1.04, "learning_rate": 3.274956873921848e-05, "loss": 1.5997, "step": 69000 }, { "epoch": 1.04, "learning_rate": 3.2624565614140354e-05, "loss": 1.6062, "step": 69500 }, { "epoch": 1.05, "learning_rate": 3.249956248906223e-05, "loss": 1.6117, "step": 70000 }, { "epoch": 1.06, "learning_rate": 3.2374559363984106e-05, "loss": 1.6268, "step": 70500 }, { "epoch": 1.07, "learning_rate": 3.224955623890598e-05, "loss": 1.601, "step": 71000 }, { "epoch": 1.07, "learning_rate": 3.212455311382785e-05, "loss": 1.5859, "step": 71500 }, { "epoch": 1.08, "learning_rate": 3.199954998874972e-05, "loss": 1.6313, "step": 72000 }, { "epoch": 1.09, "learning_rate": 3.1874546863671595e-05, "loss": 1.595, "step": 72500 }, { "epoch": 1.1, "learning_rate": 3.174954373859347e-05, "loss": 1.6144, "step": 73000 }, { "epoch": 1.1, "learning_rate": 3.162454061351534e-05, "loss": 1.5991, "step": 73500 }, { "epoch": 1.11, "learning_rate": 3.149953748843721e-05, "loss": 1.5637, "step": 74000 }, { "epoch": 1.12, "learning_rate": 3.1374534363359085e-05, "loss": 1.5957, "step": 74500 }, { "epoch": 1.13, "learning_rate": 3.124953123828096e-05, "loss": 1.588, "step": 75000 }, { "epoch": 1.13, "learning_rate": 3.1124528113202836e-05, "loss": 1.5894, "step": 75500 }, { "epoch": 1.14, "learning_rate": 3.099952498812471e-05, "loss": 1.5897, "step": 76000 }, { "epoch": 1.15, "learning_rate": 3.087452186304658e-05, "loss": 1.5835, "step": 76500 }, { "epoch": 1.16, "learning_rate": 3.074951873796845e-05, "loss": 1.6129, "step": 77000 }, { "epoch": 1.16, "learning_rate": 3.0624515612890326e-05, "loss": 1.5726, "step": 77500 }, { "epoch": 1.17, "learning_rate": 3.0499512487812198e-05, "loss": 1.5672, "step": 78000 }, { "epoch": 1.18, "learning_rate": 3.037450936273407e-05, "loss": 1.599, "step": 78500 }, { "epoch": 1.19, "learning_rate": 3.0249506237655946e-05, "loss": 1.5861, "step": 79000 }, { "epoch": 1.19, "learning_rate": 3.012450311257782e-05, "loss": 1.5977, "step": 79500 }, { "epoch": 1.2, "learning_rate": 2.999949998749969e-05, "loss": 1.5801, "step": 80000 }, { "epoch": 1.21, "learning_rate": 2.9874496862421563e-05, "loss": 1.5989, "step": 80500 }, { "epoch": 1.22, "learning_rate": 2.9749493737343435e-05, "loss": 1.5762, "step": 81000 }, { "epoch": 1.22, "learning_rate": 2.962449061226531e-05, "loss": 1.5779, "step": 81500 }, { "epoch": 1.23, "learning_rate": 2.9499487487187184e-05, "loss": 1.5988, "step": 82000 }, { "epoch": 1.24, "learning_rate": 2.9374484362109056e-05, "loss": 1.5652, "step": 82500 }, { "epoch": 1.25, "learning_rate": 2.9249481237030928e-05, "loss": 1.5817, "step": 83000 }, { "epoch": 1.25, "learning_rate": 2.91244781119528e-05, "loss": 1.535, "step": 83500 }, { "epoch": 1.26, "learning_rate": 2.8999474986874676e-05, "loss": 1.5596, "step": 84000 }, { "epoch": 1.27, "learning_rate": 2.887447186179655e-05, "loss": 1.5719, "step": 84500 }, { "epoch": 1.28, "learning_rate": 2.874946873671842e-05, "loss": 1.562, "step": 85000 }, { "epoch": 1.28, "learning_rate": 2.8624465611640293e-05, "loss": 1.5764, "step": 85500 }, { "epoch": 1.29, "learning_rate": 2.8499462486562166e-05, "loss": 1.5751, "step": 86000 }, { "epoch": 1.3, "learning_rate": 2.837445936148404e-05, "loss": 1.5826, "step": 86500 }, { "epoch": 1.31, "learning_rate": 2.8249456236405914e-05, "loss": 1.5793, "step": 87000 }, { "epoch": 1.31, "learning_rate": 2.8124453111327786e-05, "loss": 1.5526, "step": 87500 }, { "epoch": 1.32, "learning_rate": 2.799944998624966e-05, "loss": 1.5444, "step": 88000 }, { "epoch": 1.33, "learning_rate": 2.7874446861171534e-05, "loss": 1.5663, "step": 88500 }, { "epoch": 1.34, "learning_rate": 2.7749443736093407e-05, "loss": 1.5701, "step": 89000 }, { "epoch": 1.34, "learning_rate": 2.762444061101528e-05, "loss": 1.58, "step": 89500 }, { "epoch": 1.35, "learning_rate": 2.749943748593715e-05, "loss": 1.531, "step": 90000 }, { "epoch": 1.36, "learning_rate": 2.7374434360859024e-05, "loss": 1.5674, "step": 90500 }, { "epoch": 1.37, "learning_rate": 2.72494312357809e-05, "loss": 1.5475, "step": 91000 }, { "epoch": 1.37, "learning_rate": 2.7124428110702772e-05, "loss": 1.5701, "step": 91500 }, { "epoch": 1.38, "learning_rate": 2.6999424985624644e-05, "loss": 1.5722, "step": 92000 }, { "epoch": 1.39, "learning_rate": 2.6874421860546516e-05, "loss": 1.5633, "step": 92500 }, { "epoch": 1.4, "learning_rate": 2.674941873546839e-05, "loss": 1.5682, "step": 93000 }, { "epoch": 1.4, "learning_rate": 2.6624415610390265e-05, "loss": 1.5839, "step": 93500 }, { "epoch": 1.41, "learning_rate": 2.6499412485312137e-05, "loss": 1.5587, "step": 94000 }, { "epoch": 1.42, "learning_rate": 2.637440936023401e-05, "loss": 1.5585, "step": 94500 }, { "epoch": 1.43, "learning_rate": 2.624940623515588e-05, "loss": 1.5573, "step": 95000 }, { "epoch": 1.43, "learning_rate": 2.6124403110077754e-05, "loss": 1.5558, "step": 95500 }, { "epoch": 1.44, "learning_rate": 2.599939998499963e-05, "loss": 1.5818, "step": 96000 }, { "epoch": 1.45, "learning_rate": 2.5874396859921502e-05, "loss": 1.5416, "step": 96500 }, { "epoch": 1.46, "learning_rate": 2.5749393734843374e-05, "loss": 1.5566, "step": 97000 }, { "epoch": 1.46, "learning_rate": 2.5624390609765247e-05, "loss": 1.573, "step": 97500 }, { "epoch": 1.47, "learning_rate": 2.549938748468712e-05, "loss": 1.5625, "step": 98000 }, { "epoch": 1.48, "learning_rate": 2.5374384359608995e-05, "loss": 1.5688, "step": 98500 }, { "epoch": 1.49, "learning_rate": 2.5249381234530867e-05, "loss": 1.5468, "step": 99000 }, { "epoch": 1.49, "learning_rate": 2.512437810945274e-05, "loss": 1.5511, "step": 99500 }, { "epoch": 1.5, "learning_rate": 2.499937498437461e-05, "loss": 1.5325, "step": 100000 }, { "epoch": 1.51, "learning_rate": 2.4874371859296484e-05, "loss": 1.5505, "step": 100500 }, { "epoch": 1.52, "learning_rate": 2.4749368734218357e-05, "loss": 1.535, "step": 101000 }, { "epoch": 1.52, "learning_rate": 2.462436560914023e-05, "loss": 1.5471, "step": 101500 }, { "epoch": 1.53, "learning_rate": 2.44993624840621e-05, "loss": 1.5607, "step": 102000 }, { "epoch": 1.54, "learning_rate": 2.4374359358983974e-05, "loss": 1.5426, "step": 102500 }, { "epoch": 1.55, "learning_rate": 2.424935623390585e-05, "loss": 1.5488, "step": 103000 }, { "epoch": 1.55, "learning_rate": 2.4124353108827722e-05, "loss": 1.5321, "step": 103500 }, { "epoch": 1.56, "learning_rate": 2.3999349983749594e-05, "loss": 1.528, "step": 104000 }, { "epoch": 1.57, "learning_rate": 2.3874346858671467e-05, "loss": 1.5416, "step": 104500 }, { "epoch": 1.58, "learning_rate": 2.374934373359334e-05, "loss": 1.5228, "step": 105000 }, { "epoch": 1.58, "learning_rate": 2.3624340608515215e-05, "loss": 1.5114, "step": 105500 }, { "epoch": 1.59, "learning_rate": 2.3499337483437087e-05, "loss": 1.502, "step": 106000 }, { "epoch": 1.6, "learning_rate": 2.337433435835896e-05, "loss": 1.5264, "step": 106500 }, { "epoch": 1.61, "learning_rate": 2.324933123328083e-05, "loss": 1.5065, "step": 107000 }, { "epoch": 1.61, "learning_rate": 2.3124328108202707e-05, "loss": 1.525, "step": 107500 }, { "epoch": 1.62, "learning_rate": 2.299932498312458e-05, "loss": 1.5159, "step": 108000 }, { "epoch": 1.63, "learning_rate": 2.2874321858046452e-05, "loss": 1.521, "step": 108500 }, { "epoch": 1.64, "learning_rate": 2.2749318732968324e-05, "loss": 1.5116, "step": 109000 }, { "epoch": 1.64, "learning_rate": 2.2624315607890197e-05, "loss": 1.5392, "step": 109500 }, { "epoch": 1.65, "learning_rate": 2.2499312482812073e-05, "loss": 1.5245, "step": 110000 }, { "epoch": 1.66, "learning_rate": 2.2374309357733945e-05, "loss": 1.5392, "step": 110500 }, { "epoch": 1.67, "learning_rate": 2.2249306232655817e-05, "loss": 1.5114, "step": 111000 }, { "epoch": 1.67, "learning_rate": 2.212430310757769e-05, "loss": 1.55, "step": 111500 }, { "epoch": 1.68, "learning_rate": 2.1999299982499562e-05, "loss": 1.487, "step": 112000 }, { "epoch": 1.69, "learning_rate": 2.1874296857421438e-05, "loss": 1.5393, "step": 112500 }, { "epoch": 1.7, "learning_rate": 2.174929373234331e-05, "loss": 1.529, "step": 113000 }, { "epoch": 1.7, "learning_rate": 2.1624290607265182e-05, "loss": 1.5167, "step": 113500 }, { "epoch": 1.71, "learning_rate": 2.1499287482187055e-05, "loss": 1.5101, "step": 114000 }, { "epoch": 1.72, "learning_rate": 2.1374284357108927e-05, "loss": 1.5113, "step": 114500 }, { "epoch": 1.73, "learning_rate": 2.1249281232030803e-05, "loss": 1.5353, "step": 115000 }, { "epoch": 1.73, "learning_rate": 2.1124278106952675e-05, "loss": 1.4933, "step": 115500 }, { "epoch": 1.74, "learning_rate": 2.0999274981874548e-05, "loss": 1.5104, "step": 116000 }, { "epoch": 1.75, "learning_rate": 2.087427185679642e-05, "loss": 1.4918, "step": 116500 }, { "epoch": 1.76, "learning_rate": 2.0749268731718292e-05, "loss": 1.518, "step": 117000 }, { "epoch": 1.76, "learning_rate": 2.0624265606640168e-05, "loss": 1.4962, "step": 117500 }, { "epoch": 1.77, "learning_rate": 2.049926248156204e-05, "loss": 1.5395, "step": 118000 }, { "epoch": 1.78, "learning_rate": 2.0374259356483913e-05, "loss": 1.5027, "step": 118500 }, { "epoch": 1.79, "learning_rate": 2.0249256231405785e-05, "loss": 1.5285, "step": 119000 }, { "epoch": 1.79, "learning_rate": 2.0124253106327657e-05, "loss": 1.5141, "step": 119500 }, { "epoch": 1.8, "learning_rate": 1.9999249981249533e-05, "loss": 1.4661, "step": 120000 }, { "epoch": 1.81, "learning_rate": 1.9874246856171406e-05, "loss": 1.5139, "step": 120500 }, { "epoch": 1.82, "learning_rate": 1.9749243731093278e-05, "loss": 1.4976, "step": 121000 }, { "epoch": 1.82, "learning_rate": 1.962424060601515e-05, "loss": 1.4833, "step": 121500 }, { "epoch": 1.83, "learning_rate": 1.9499237480937026e-05, "loss": 1.4846, "step": 122000 }, { "epoch": 1.84, "learning_rate": 1.93742343558589e-05, "loss": 1.4874, "step": 122500 }, { "epoch": 1.85, "learning_rate": 1.924923123078077e-05, "loss": 1.4739, "step": 123000 }, { "epoch": 1.85, "learning_rate": 1.9124228105702643e-05, "loss": 1.502, "step": 123500 }, { "epoch": 1.86, "learning_rate": 1.8999224980624515e-05, "loss": 1.4991, "step": 124000 }, { "epoch": 1.87, "learning_rate": 1.887422185554639e-05, "loss": 1.4695, "step": 124500 }, { "epoch": 1.88, "learning_rate": 1.8749218730468263e-05, "loss": 1.5081, "step": 125000 }, { "epoch": 1.88, "learning_rate": 1.8624215605390136e-05, "loss": 1.4977, "step": 125500 }, { "epoch": 1.89, "learning_rate": 1.8499212480312008e-05, "loss": 1.4885, "step": 126000 }, { "epoch": 1.9, "learning_rate": 1.837420935523388e-05, "loss": 1.4812, "step": 126500 }, { "epoch": 1.91, "learning_rate": 1.8249206230155756e-05, "loss": 1.4861, "step": 127000 }, { "epoch": 1.91, "learning_rate": 1.812420310507763e-05, "loss": 1.4895, "step": 127500 }, { "epoch": 1.92, "learning_rate": 1.79991999799995e-05, "loss": 1.4823, "step": 128000 }, { "epoch": 1.93, "learning_rate": 1.7874196854921373e-05, "loss": 1.4821, "step": 128500 }, { "epoch": 1.94, "learning_rate": 1.7749193729843246e-05, "loss": 1.5037, "step": 129000 }, { "epoch": 1.94, "learning_rate": 1.762419060476512e-05, "loss": 1.4783, "step": 129500 }, { "epoch": 1.95, "learning_rate": 1.7499187479686994e-05, "loss": 1.5003, "step": 130000 }, { "epoch": 1.96, "learning_rate": 1.7374184354608866e-05, "loss": 1.4846, "step": 130500 }, { "epoch": 1.97, "learning_rate": 1.724918122953074e-05, "loss": 1.4976, "step": 131000 }, { "epoch": 1.97, "learning_rate": 1.712417810445261e-05, "loss": 1.4989, "step": 131500 }, { "epoch": 1.98, "learning_rate": 1.6999174979374487e-05, "loss": 1.4746, "step": 132000 }, { "epoch": 1.99, "learning_rate": 1.687417185429636e-05, "loss": 1.4871, "step": 132500 }, { "epoch": 2.0, "learning_rate": 1.674916872921823e-05, "loss": 1.494, "step": 133000 }, { "epoch": 2.0, "learning_rate": 1.6624165604140104e-05, "loss": 1.4163, "step": 133500 }, { "epoch": 2.01, "learning_rate": 1.6499162479061976e-05, "loss": 1.3369, "step": 134000 }, { "epoch": 2.02, "learning_rate": 1.6374159353983852e-05, "loss": 1.3468, "step": 134500 }, { "epoch": 2.03, "learning_rate": 1.6249156228905724e-05, "loss": 1.3477, "step": 135000 }, { "epoch": 2.03, "learning_rate": 1.6124153103827596e-05, "loss": 1.3267, "step": 135500 }, { "epoch": 2.04, "learning_rate": 1.599914997874947e-05, "loss": 1.3454, "step": 136000 }, { "epoch": 2.05, "learning_rate": 1.587414685367134e-05, "loss": 1.3346, "step": 136500 }, { "epoch": 2.06, "learning_rate": 1.5749143728593217e-05, "loss": 1.3609, "step": 137000 }, { "epoch": 2.06, "learning_rate": 1.562414060351509e-05, "loss": 1.3203, "step": 137500 }, { "epoch": 2.07, "learning_rate": 1.549913747843696e-05, "loss": 1.3446, "step": 138000 }, { "epoch": 2.08, "learning_rate": 1.5374134353358834e-05, "loss": 1.3297, "step": 138500 }, { "epoch": 2.09, "learning_rate": 1.5249131228280708e-05, "loss": 1.3478, "step": 139000 }, { "epoch": 2.09, "learning_rate": 1.512412810320258e-05, "loss": 1.3475, "step": 139500 }, { "epoch": 2.1, "learning_rate": 1.4999124978124454e-05, "loss": 1.2979, "step": 140000 }, { "epoch": 2.11, "learning_rate": 1.4874121853046327e-05, "loss": 1.3368, "step": 140500 }, { "epoch": 2.12, "learning_rate": 1.47491187279682e-05, "loss": 1.3378, "step": 141000 }, { "epoch": 2.12, "learning_rate": 1.4624115602890073e-05, "loss": 1.3591, "step": 141500 }, { "epoch": 2.13, "learning_rate": 1.4499112477811947e-05, "loss": 1.3627, "step": 142000 }, { "epoch": 2.14, "learning_rate": 1.437410935273382e-05, "loss": 1.3245, "step": 142500 }, { "epoch": 2.15, "learning_rate": 1.4249106227655692e-05, "loss": 1.3322, "step": 143000 }, { "epoch": 2.15, "learning_rate": 1.4124103102577566e-05, "loss": 1.3387, "step": 143500 }, { "epoch": 2.16, "learning_rate": 1.3999099977499438e-05, "loss": 1.3613, "step": 144000 }, { "epoch": 2.17, "learning_rate": 1.3874096852421312e-05, "loss": 1.3205, "step": 144500 }, { "epoch": 2.18, "learning_rate": 1.3749093727343185e-05, "loss": 1.3701, "step": 145000 }, { "epoch": 2.18, "learning_rate": 1.3624090602265057e-05, "loss": 1.3483, "step": 145500 }, { "epoch": 2.19, "learning_rate": 1.3499087477186931e-05, "loss": 1.3294, "step": 146000 }, { "epoch": 2.2, "learning_rate": 1.3374084352108803e-05, "loss": 1.3255, "step": 146500 }, { "epoch": 2.21, "learning_rate": 1.3249081227030678e-05, "loss": 1.3477, "step": 147000 }, { "epoch": 2.21, "learning_rate": 1.312407810195255e-05, "loss": 1.3608, "step": 147500 }, { "epoch": 2.22, "learning_rate": 1.2999074976874422e-05, "loss": 1.3364, "step": 148000 }, { "epoch": 2.23, "learning_rate": 1.2874071851796296e-05, "loss": 1.3534, "step": 148500 }, { "epoch": 2.24, "learning_rate": 1.2749068726718169e-05, "loss": 1.347, "step": 149000 }, { "epoch": 2.24, "learning_rate": 1.2624065601640043e-05, "loss": 1.3482, "step": 149500 }, { "epoch": 2.25, "learning_rate": 1.2499062476561915e-05, "loss": 1.356, "step": 150000 }, { "epoch": 2.26, "learning_rate": 1.2374059351483789e-05, "loss": 1.336, "step": 150500 }, { "epoch": 2.27, "learning_rate": 1.2249056226405661e-05, "loss": 1.3676, "step": 151000 }, { "epoch": 2.27, "learning_rate": 1.2124053101327534e-05, "loss": 1.3444, "step": 151500 }, { "epoch": 2.28, "learning_rate": 1.1999049976249408e-05, "loss": 1.3574, "step": 152000 }, { "epoch": 2.29, "learning_rate": 1.187404685117128e-05, "loss": 1.3093, "step": 152500 }, { "epoch": 2.3, "learning_rate": 1.1749043726093154e-05, "loss": 1.3318, "step": 153000 }, { "epoch": 2.3, "learning_rate": 1.1624040601015027e-05, "loss": 1.3291, "step": 153500 }, { "epoch": 2.31, "learning_rate": 1.1499037475936899e-05, "loss": 1.3255, "step": 154000 }, { "epoch": 2.32, "learning_rate": 1.1374034350858773e-05, "loss": 1.3142, "step": 154500 }, { "epoch": 2.33, "learning_rate": 1.1249031225780645e-05, "loss": 1.3293, "step": 155000 }, { "epoch": 2.33, "learning_rate": 1.112402810070252e-05, "loss": 1.342, "step": 155500 }, { "epoch": 2.34, "learning_rate": 1.0999024975624392e-05, "loss": 1.308, "step": 156000 }, { "epoch": 2.35, "learning_rate": 1.0874021850546264e-05, "loss": 1.3136, "step": 156500 }, { "epoch": 2.36, "learning_rate": 1.0749018725468136e-05, "loss": 1.3274, "step": 157000 }, { "epoch": 2.36, "learning_rate": 1.0624015600390009e-05, "loss": 1.3306, "step": 157500 }, { "epoch": 2.37, "learning_rate": 1.0499012475311883e-05, "loss": 1.3347, "step": 158000 }, { "epoch": 2.38, "learning_rate": 1.0374009350233755e-05, "loss": 1.3135, "step": 158500 }, { "epoch": 2.39, "learning_rate": 1.024900622515563e-05, "loss": 1.3239, "step": 159000 }, { "epoch": 2.39, "learning_rate": 1.0124003100077502e-05, "loss": 1.3085, "step": 159500 }, { "epoch": 2.4, "learning_rate": 9.998999974999376e-06, "loss": 1.3323, "step": 160000 }, { "epoch": 2.41, "learning_rate": 9.873996849921248e-06, "loss": 1.3372, "step": 160500 }, { "epoch": 2.42, "learning_rate": 9.74899372484312e-06, "loss": 1.317, "step": 161000 }, { "epoch": 2.42, "learning_rate": 9.623990599764994e-06, "loss": 1.3273, "step": 161500 }, { "epoch": 2.43, "learning_rate": 9.498987474686867e-06, "loss": 1.3139, "step": 162000 }, { "epoch": 2.44, "learning_rate": 9.37398434960874e-06, "loss": 1.3027, "step": 162500 }, { "epoch": 2.45, "learning_rate": 9.248981224530613e-06, "loss": 1.3, "step": 163000 }, { "epoch": 2.45, "learning_rate": 9.123978099452486e-06, "loss": 1.3465, "step": 163500 }, { "epoch": 2.46, "learning_rate": 8.99897497437436e-06, "loss": 1.3039, "step": 164000 }, { "epoch": 2.47, "learning_rate": 8.873971849296232e-06, "loss": 1.3188, "step": 164500 }, { "epoch": 2.48, "learning_rate": 8.748968724218106e-06, "loss": 1.3457, "step": 165000 }, { "epoch": 2.48, "learning_rate": 8.623965599139978e-06, "loss": 1.3377, "step": 165500 }, { "epoch": 2.49, "learning_rate": 8.49896247406185e-06, "loss": 1.3137, "step": 166000 }, { "epoch": 2.5, "learning_rate": 8.373959348983725e-06, "loss": 1.3265, "step": 166500 }, { "epoch": 2.51, "learning_rate": 8.248956223905597e-06, "loss": 1.3429, "step": 167000 }, { "epoch": 2.51, "learning_rate": 8.123953098827471e-06, "loss": 1.3233, "step": 167500 }, { "epoch": 2.52, "learning_rate": 7.998949973749343e-06, "loss": 1.2967, "step": 168000 }, { "epoch": 2.53, "learning_rate": 7.873946848671218e-06, "loss": 1.3232, "step": 168500 }, { "epoch": 2.54, "learning_rate": 7.74894372359309e-06, "loss": 1.2925, "step": 169000 }, { "epoch": 2.54, "learning_rate": 7.623940598514963e-06, "loss": 1.3128, "step": 169500 }, { "epoch": 2.55, "learning_rate": 7.498937473436836e-06, "loss": 1.3273, "step": 170000 }, { "epoch": 2.56, "learning_rate": 7.373934348358709e-06, "loss": 1.3101, "step": 170500 }, { "epoch": 2.57, "learning_rate": 7.248931223280582e-06, "loss": 1.3044, "step": 171000 }, { "epoch": 2.57, "learning_rate": 7.123928098202455e-06, "loss": 1.2993, "step": 171500 }, { "epoch": 2.58, "learning_rate": 6.998924973124328e-06, "loss": 1.3313, "step": 172000 }, { "epoch": 2.59, "learning_rate": 6.873921848046201e-06, "loss": 1.3018, "step": 172500 }, { "epoch": 2.6, "learning_rate": 6.748918722968075e-06, "loss": 1.3022, "step": 173000 }, { "epoch": 2.6, "learning_rate": 6.623915597889947e-06, "loss": 1.2963, "step": 173500 }, { "epoch": 2.61, "learning_rate": 6.49891247281182e-06, "loss": 1.3185, "step": 174000 }, { "epoch": 2.62, "learning_rate": 6.373909347733693e-06, "loss": 1.3228, "step": 174500 }, { "epoch": 2.63, "learning_rate": 6.2489062226555666e-06, "loss": 1.3196, "step": 175000 }, { "epoch": 2.63, "learning_rate": 6.12390309757744e-06, "loss": 1.3052, "step": 175500 }, { "epoch": 2.64, "learning_rate": 5.998899972499312e-06, "loss": 1.2999, "step": 176000 }, { "epoch": 2.65, "learning_rate": 5.873896847421185e-06, "loss": 1.3266, "step": 176500 }, { "epoch": 2.66, "learning_rate": 5.7488937223430585e-06, "loss": 1.3008, "step": 177000 }, { "epoch": 2.66, "learning_rate": 5.623890597264932e-06, "loss": 1.316, "step": 177500 }, { "epoch": 2.67, "learning_rate": 5.498887472186805e-06, "loss": 1.3147, "step": 178000 }, { "epoch": 2.68, "learning_rate": 5.373884347108678e-06, "loss": 1.3141, "step": 178500 }, { "epoch": 2.69, "learning_rate": 5.2488812220305505e-06, "loss": 1.3112, "step": 179000 }, { "epoch": 2.69, "learning_rate": 5.123878096952424e-06, "loss": 1.3116, "step": 179500 }, { "epoch": 2.7, "learning_rate": 4.998874971874297e-06, "loss": 1.3018, "step": 180000 }, { "epoch": 2.71, "learning_rate": 4.87387184679617e-06, "loss": 1.3224, "step": 180500 }, { "epoch": 2.72, "learning_rate": 4.748868721718043e-06, "loss": 1.2892, "step": 181000 }, { "epoch": 2.72, "learning_rate": 4.6238655966399165e-06, "loss": 1.2784, "step": 181500 }, { "epoch": 2.73, "learning_rate": 4.498862471561789e-06, "loss": 1.3124, "step": 182000 }, { "epoch": 2.74, "learning_rate": 4.373859346483662e-06, "loss": 1.3137, "step": 182500 }, { "epoch": 2.75, "learning_rate": 4.248856221405535e-06, "loss": 1.3028, "step": 183000 }, { "epoch": 2.75, "learning_rate": 4.1238530963274084e-06, "loss": 1.3232, "step": 183500 }, { "epoch": 2.76, "learning_rate": 3.998849971249282e-06, "loss": 1.2967, "step": 184000 }, { "epoch": 2.77, "learning_rate": 3.873846846171154e-06, "loss": 1.285, "step": 184500 }, { "epoch": 2.78, "learning_rate": 3.7488437210930276e-06, "loss": 1.2919, "step": 185000 }, { "epoch": 2.78, "learning_rate": 3.6238405960149004e-06, "loss": 1.3022, "step": 185500 }, { "epoch": 2.79, "learning_rate": 3.4988374709367736e-06, "loss": 1.2914, "step": 186000 }, { "epoch": 2.8, "learning_rate": 3.373834345858647e-06, "loss": 1.3033, "step": 186500 }, { "epoch": 2.81, "learning_rate": 3.2488312207805196e-06, "loss": 1.2823, "step": 187000 }, { "epoch": 2.81, "learning_rate": 3.1238280957023928e-06, "loss": 1.3115, "step": 187500 }, { "epoch": 2.82, "learning_rate": 2.998824970624266e-06, "loss": 1.2941, "step": 188000 }, { "epoch": 2.83, "learning_rate": 2.8738218455461387e-06, "loss": 1.2785, "step": 188500 }, { "epoch": 2.84, "learning_rate": 2.748818720468012e-06, "loss": 1.2913, "step": 189000 }, { "epoch": 2.84, "learning_rate": 2.6238155953898847e-06, "loss": 1.3162, "step": 189500 }, { "epoch": 2.85, "learning_rate": 2.498812470311758e-06, "loss": 1.3114, "step": 190000 }, { "epoch": 2.86, "learning_rate": 2.373809345233631e-06, "loss": 1.2898, "step": 190500 }, { "epoch": 2.87, "learning_rate": 2.248806220155504e-06, "loss": 1.313, "step": 191000 }, { "epoch": 2.87, "learning_rate": 2.123803095077377e-06, "loss": 1.2896, "step": 191500 }, { "epoch": 2.88, "learning_rate": 1.9987999699992503e-06, "loss": 1.3174, "step": 192000 }, { "epoch": 2.89, "learning_rate": 1.8737968449211233e-06, "loss": 1.2951, "step": 192500 }, { "epoch": 2.9, "learning_rate": 1.7487937198429963e-06, "loss": 1.3192, "step": 193000 }, { "epoch": 2.9, "learning_rate": 1.6237905947648693e-06, "loss": 1.3053, "step": 193500 }, { "epoch": 2.91, "learning_rate": 1.4987874696867423e-06, "loss": 1.3045, "step": 194000 }, { "epoch": 2.92, "learning_rate": 1.3737843446086152e-06, "loss": 1.2992, "step": 194500 }, { "epoch": 2.93, "learning_rate": 1.2487812195304882e-06, "loss": 1.2643, "step": 195000 }, { "epoch": 2.93, "learning_rate": 1.1237780944523614e-06, "loss": 1.2917, "step": 195500 }, { "epoch": 2.94, "learning_rate": 9.987749693742344e-07, "loss": 1.294, "step": 196000 }, { "epoch": 2.95, "learning_rate": 8.737718442961075e-07, "loss": 1.289, "step": 196500 }, { "epoch": 2.96, "learning_rate": 7.487687192179805e-07, "loss": 1.2809, "step": 197000 }, { "epoch": 2.96, "learning_rate": 6.237655941398535e-07, "loss": 1.3175, "step": 197500 }, { "epoch": 2.97, "learning_rate": 4.987624690617266e-07, "loss": 1.3111, "step": 198000 }, { "epoch": 2.98, "learning_rate": 3.7375934398359957e-07, "loss": 1.3019, "step": 198500 }, { "epoch": 2.99, "learning_rate": 2.4875621890547267e-07, "loss": 1.3065, "step": 199000 }, { "epoch": 2.99, "learning_rate": 1.237530938273457e-07, "loss": 1.2757, "step": 199500 }, { "epoch": 3.0, "step": 199995, "total_flos": 3.01955982127104e+17, "train_loss": 1.6198288996531554, "train_runtime": 56046.7089, "train_samples_per_second": 35.683, "train_steps_per_second": 3.568 } ], "max_steps": 199995, "num_train_epochs": 3, "total_flos": 3.01955982127104e+17, "trial_name": null, "trial_params": null }