{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.896, "global_step": 280000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 3.868455096e-05, "loss": 0.0719, "step": 500 }, { "epoch": 0.0, "learning_rate": 3.866907096e-05, "loss": 0.0693, "step": 1000 }, { "epoch": 0.0, "learning_rate": 3.865359096e-05, "loss": 0.0708, "step": 1500 }, { "epoch": 0.01, "learning_rate": 3.8638110959999996e-05, "loss": 0.0728, "step": 2000 }, { "epoch": 0.01, "learning_rate": 3.862263096e-05, "loss": 0.0701, "step": 2500 }, { "epoch": 0.01, "learning_rate": 3.860715096e-05, "loss": 0.0707, "step": 3000 }, { "epoch": 0.01, "learning_rate": 3.859167096e-05, "loss": 0.0704, "step": 3500 }, { "epoch": 0.01, "learning_rate": 3.857619096e-05, "loss": 0.0707, "step": 4000 }, { "epoch": 0.01, "learning_rate": 3.8560710959999996e-05, "loss": 0.0688, "step": 4500 }, { "epoch": 0.02, "learning_rate": 3.854523096e-05, "loss": 0.0703, "step": 5000 }, { "epoch": 0.02, "learning_rate": 3.852975096e-05, "loss": 0.0683, "step": 5500 }, { "epoch": 0.02, "learning_rate": 3.851427096e-05, "loss": 0.0697, "step": 6000 }, { "epoch": 0.02, "learning_rate": 3.849879096e-05, "loss": 0.07, "step": 6500 }, { "epoch": 0.02, "learning_rate": 3.8483310959999995e-05, "loss": 0.0682, "step": 7000 }, { "epoch": 0.02, "learning_rate": 3.846783096e-05, "loss": 0.0695, "step": 7500 }, { "epoch": 0.03, "learning_rate": 3.845235096e-05, "loss": 0.0712, "step": 8000 }, { "epoch": 0.03, "learning_rate": 3.843687096e-05, "loss": 0.0694, "step": 8500 }, { "epoch": 0.03, "learning_rate": 3.842139096e-05, "loss": 0.0687, "step": 9000 }, { "epoch": 0.03, "learning_rate": 3.8405910959999995e-05, "loss": 0.0716, "step": 9500 }, { "epoch": 0.03, "learning_rate": 3.8390430959999996e-05, "loss": 0.0691, "step": 10000 }, { "epoch": 0.03, "learning_rate": 3.837495096e-05, "loss": 0.0701, "step": 10500 }, { "epoch": 0.04, "learning_rate": 3.835947096e-05, "loss": 0.0707, "step": 11000 }, { "epoch": 0.04, "learning_rate": 3.834399096e-05, "loss": 0.0702, "step": 11500 }, { "epoch": 0.04, "learning_rate": 3.8328510959999995e-05, "loss": 0.0694, "step": 12000 }, { "epoch": 0.04, "learning_rate": 3.8313030959999996e-05, "loss": 0.0706, "step": 12500 }, { "epoch": 0.04, "learning_rate": 3.829755096e-05, "loss": 0.07, "step": 13000 }, { "epoch": 0.04, "learning_rate": 3.828207096e-05, "loss": 0.0695, "step": 13500 }, { "epoch": 0.04, "learning_rate": 3.826659096e-05, "loss": 0.0688, "step": 14000 }, { "epoch": 0.05, "learning_rate": 3.8251110959999994e-05, "loss": 0.0683, "step": 14500 }, { "epoch": 0.05, "learning_rate": 3.823563096e-05, "loss": 0.0696, "step": 15000 }, { "epoch": 0.05, "learning_rate": 3.822015096e-05, "loss": 0.0703, "step": 15500 }, { "epoch": 0.05, "learning_rate": 3.820467096e-05, "loss": 0.0715, "step": 16000 }, { "epoch": 0.05, "learning_rate": 3.818919096e-05, "loss": 0.0688, "step": 16500 }, { "epoch": 0.05, "learning_rate": 3.8173710959999994e-05, "loss": 0.0695, "step": 17000 }, { "epoch": 0.06, "learning_rate": 3.815823096e-05, "loss": 0.0711, "step": 17500 }, { "epoch": 0.06, "learning_rate": 3.8142750960000003e-05, "loss": 0.0692, "step": 18000 }, { "epoch": 0.06, "learning_rate": 3.812727096e-05, "loss": 0.069, "step": 18500 }, { "epoch": 0.06, "learning_rate": 3.811179096e-05, "loss": 0.0689, "step": 19000 }, { "epoch": 0.06, "learning_rate": 3.809631096e-05, "loss": 0.0698, "step": 19500 }, { "epoch": 0.06, "learning_rate": 3.808083096e-05, "loss": 0.0723, "step": 20000 }, { "epoch": 0.07, "learning_rate": 3.806535096e-05, "loss": 0.0682, "step": 20500 }, { "epoch": 0.07, "learning_rate": 3.804987096e-05, "loss": 0.0686, "step": 21000 }, { "epoch": 0.07, "learning_rate": 3.803439096e-05, "loss": 0.0719, "step": 21500 }, { "epoch": 0.07, "learning_rate": 3.801891096e-05, "loss": 0.0685, "step": 22000 }, { "epoch": 0.07, "learning_rate": 3.800343096e-05, "loss": 0.0692, "step": 22500 }, { "epoch": 0.07, "learning_rate": 3.798795096e-05, "loss": 0.0697, "step": 23000 }, { "epoch": 0.08, "learning_rate": 3.797247096e-05, "loss": 0.0711, "step": 23500 }, { "epoch": 0.08, "learning_rate": 3.795699096e-05, "loss": 0.0687, "step": 24000 }, { "epoch": 0.08, "learning_rate": 3.794151096e-05, "loss": 0.0693, "step": 24500 }, { "epoch": 0.08, "learning_rate": 3.792603096e-05, "loss": 0.069, "step": 25000 }, { "epoch": 0.08, "learning_rate": 3.791055096e-05, "loss": 0.071, "step": 25500 }, { "epoch": 0.08, "learning_rate": 3.789507096e-05, "loss": 0.0692, "step": 26000 }, { "epoch": 0.08, "learning_rate": 3.787959096e-05, "loss": 0.07, "step": 26500 }, { "epoch": 0.09, "learning_rate": 3.786411096e-05, "loss": 0.07, "step": 27000 }, { "epoch": 0.09, "learning_rate": 3.784863096e-05, "loss": 0.0703, "step": 27500 }, { "epoch": 0.09, "learning_rate": 3.783315096e-05, "loss": 0.0684, "step": 28000 }, { "epoch": 0.09, "learning_rate": 3.781767096e-05, "loss": 0.0704, "step": 28500 }, { "epoch": 0.09, "learning_rate": 3.780219096e-05, "loss": 0.0692, "step": 29000 }, { "epoch": 0.09, "learning_rate": 3.778671096e-05, "loss": 0.0699, "step": 29500 }, { "epoch": 0.1, "learning_rate": 3.777123096e-05, "loss": 0.0692, "step": 30000 }, { "epoch": 0.1, "learning_rate": 3.775575096e-05, "loss": 0.0686, "step": 30500 }, { "epoch": 0.1, "learning_rate": 3.7740270959999996e-05, "loss": 0.0694, "step": 31000 }, { "epoch": 0.1, "learning_rate": 3.772479096e-05, "loss": 0.0707, "step": 31500 }, { "epoch": 0.1, "learning_rate": 3.770931096e-05, "loss": 0.0682, "step": 32000 }, { "epoch": 0.1, "learning_rate": 3.769383096e-05, "loss": 0.069, "step": 32500 }, { "epoch": 0.11, "learning_rate": 3.767835096e-05, "loss": 0.0705, "step": 33000 }, { "epoch": 0.11, "learning_rate": 3.7662870959999996e-05, "loss": 0.0681, "step": 33500 }, { "epoch": 0.11, "learning_rate": 3.764739096e-05, "loss": 0.0672, "step": 34000 }, { "epoch": 0.11, "learning_rate": 3.763191096e-05, "loss": 0.0694, "step": 34500 }, { "epoch": 0.11, "learning_rate": 3.761643096e-05, "loss": 0.0676, "step": 35000 }, { "epoch": 0.11, "learning_rate": 3.760095096e-05, "loss": 0.0677, "step": 35500 }, { "epoch": 0.12, "learning_rate": 3.7585470959999996e-05, "loss": 0.0691, "step": 36000 }, { "epoch": 0.12, "learning_rate": 3.756999096e-05, "loss": 0.0682, "step": 36500 }, { "epoch": 0.12, "learning_rate": 3.755451096e-05, "loss": 0.067, "step": 37000 }, { "epoch": 0.12, "learning_rate": 3.753903096e-05, "loss": 0.0666, "step": 37500 }, { "epoch": 0.12, "learning_rate": 3.752355096e-05, "loss": 0.0701, "step": 38000 }, { "epoch": 0.12, "learning_rate": 3.7508070959999996e-05, "loss": 0.0685, "step": 38500 }, { "epoch": 0.12, "learning_rate": 3.749259096e-05, "loss": 0.0696, "step": 39000 }, { "epoch": 0.13, "learning_rate": 3.747711096e-05, "loss": 0.0672, "step": 39500 }, { "epoch": 0.13, "learning_rate": 3.746163096e-05, "loss": 0.0678, "step": 40000 }, { "epoch": 0.13, "learning_rate": 3.744615096e-05, "loss": 0.0686, "step": 40500 }, { "epoch": 0.13, "learning_rate": 3.7430670959999995e-05, "loss": 0.0684, "step": 41000 }, { "epoch": 0.13, "learning_rate": 3.7415190959999997e-05, "loss": 0.069, "step": 41500 }, { "epoch": 0.13, "learning_rate": 3.739971096e-05, "loss": 0.066, "step": 42000 }, { "epoch": 0.14, "learning_rate": 3.738423096e-05, "loss": 0.0681, "step": 42500 }, { "epoch": 0.14, "learning_rate": 3.736875096e-05, "loss": 0.0664, "step": 43000 }, { "epoch": 0.14, "learning_rate": 3.7353270959999995e-05, "loss": 0.0677, "step": 43500 }, { "epoch": 0.14, "learning_rate": 3.7337790959999996e-05, "loss": 0.0696, "step": 44000 }, { "epoch": 0.14, "learning_rate": 3.732231096e-05, "loss": 0.0656, "step": 44500 }, { "epoch": 0.14, "learning_rate": 3.730683096e-05, "loss": 0.0709, "step": 45000 }, { "epoch": 0.15, "learning_rate": 3.729135096e-05, "loss": 0.0668, "step": 45500 }, { "epoch": 0.15, "learning_rate": 3.7275870959999995e-05, "loss": 0.0694, "step": 46000 }, { "epoch": 0.15, "learning_rate": 3.726039096e-05, "loss": 0.0696, "step": 46500 }, { "epoch": 0.15, "learning_rate": 3.724491096e-05, "loss": 0.0685, "step": 47000 }, { "epoch": 0.15, "learning_rate": 3.722943096e-05, "loss": 0.067, "step": 47500 }, { "epoch": 0.15, "learning_rate": 3.721395096e-05, "loss": 0.0678, "step": 48000 }, { "epoch": 0.16, "learning_rate": 3.7198470959999994e-05, "loss": 0.0673, "step": 48500 }, { "epoch": 0.16, "learning_rate": 3.718299096e-05, "loss": 0.0668, "step": 49000 }, { "epoch": 0.16, "learning_rate": 3.7167510960000004e-05, "loss": 0.0665, "step": 49500 }, { "epoch": 0.16, "learning_rate": 3.715203096e-05, "loss": 0.067, "step": 50000 }, { "epoch": 0.16, "learning_rate": 3.713655096e-05, "loss": 0.0675, "step": 50500 }, { "epoch": 0.16, "learning_rate": 3.712107096e-05, "loss": 0.0665, "step": 51000 }, { "epoch": 0.16, "learning_rate": 3.710559096e-05, "loss": 0.0681, "step": 51500 }, { "epoch": 0.17, "learning_rate": 3.709011096e-05, "loss": 0.0699, "step": 52000 }, { "epoch": 0.17, "learning_rate": 3.707463096e-05, "loss": 0.0663, "step": 52500 }, { "epoch": 0.17, "learning_rate": 3.705915096e-05, "loss": 0.0672, "step": 53000 }, { "epoch": 0.17, "learning_rate": 3.704367096e-05, "loss": 0.0682, "step": 53500 }, { "epoch": 0.17, "learning_rate": 3.702819096e-05, "loss": 0.0672, "step": 54000 }, { "epoch": 0.17, "learning_rate": 3.701271096e-05, "loss": 0.0677, "step": 54500 }, { "epoch": 0.18, "learning_rate": 3.699723096e-05, "loss": 0.066, "step": 55000 }, { "epoch": 0.18, "learning_rate": 3.698175096e-05, "loss": 0.067, "step": 55500 }, { "epoch": 0.18, "learning_rate": 3.696627096e-05, "loss": 0.0678, "step": 56000 }, { "epoch": 0.18, "learning_rate": 3.695079096e-05, "loss": 0.0671, "step": 56500 }, { "epoch": 0.18, "learning_rate": 3.693531096e-05, "loss": 0.0669, "step": 57000 }, { "epoch": 0.18, "learning_rate": 3.691983096e-05, "loss": 0.0702, "step": 57500 }, { "epoch": 0.19, "learning_rate": 3.690435096e-05, "loss": 0.0692, "step": 58000 }, { "epoch": 0.19, "learning_rate": 3.688887096e-05, "loss": 0.0672, "step": 58500 }, { "epoch": 0.19, "learning_rate": 3.687339096e-05, "loss": 0.0675, "step": 59000 }, { "epoch": 0.19, "learning_rate": 3.685791096e-05, "loss": 0.0665, "step": 59500 }, { "epoch": 0.19, "learning_rate": 3.684243096e-05, "loss": 0.0683, "step": 60000 }, { "epoch": 0.19, "learning_rate": 3.682695096e-05, "loss": 0.0668, "step": 60500 }, { "epoch": 0.2, "learning_rate": 3.681147096e-05, "loss": 0.0675, "step": 61000 }, { "epoch": 0.2, "learning_rate": 3.679599096e-05, "loss": 0.0661, "step": 61500 }, { "epoch": 0.2, "learning_rate": 3.678051096e-05, "loss": 0.0688, "step": 62000 }, { "epoch": 0.2, "learning_rate": 3.676503096e-05, "loss": 0.0695, "step": 62500 }, { "epoch": 0.2, "learning_rate": 3.674955096e-05, "loss": 0.0694, "step": 63000 }, { "epoch": 0.2, "learning_rate": 3.673407096e-05, "loss": 0.0702, "step": 63500 }, { "epoch": 0.2, "learning_rate": 3.671859096e-05, "loss": 0.0679, "step": 64000 }, { "epoch": 0.21, "learning_rate": 3.670311096e-05, "loss": 0.0671, "step": 64500 }, { "epoch": 0.21, "learning_rate": 3.6687630959999996e-05, "loss": 0.0671, "step": 65000 }, { "epoch": 0.21, "learning_rate": 3.667215096e-05, "loss": 0.0671, "step": 65500 }, { "epoch": 0.21, "learning_rate": 3.665667096e-05, "loss": 0.0679, "step": 66000 }, { "epoch": 0.21, "learning_rate": 3.664119096e-05, "loss": 0.0681, "step": 66500 }, { "epoch": 0.21, "learning_rate": 3.662571096e-05, "loss": 0.0672, "step": 67000 }, { "epoch": 0.22, "learning_rate": 3.6610230959999996e-05, "loss": 0.0678, "step": 67500 }, { "epoch": 0.22, "learning_rate": 3.659475096e-05, "loss": 0.0669, "step": 68000 }, { "epoch": 0.22, "learning_rate": 3.657927096e-05, "loss": 0.068, "step": 68500 }, { "epoch": 0.22, "learning_rate": 3.656379096e-05, "loss": 0.0674, "step": 69000 }, { "epoch": 0.22, "learning_rate": 3.654831096e-05, "loss": 0.0663, "step": 69500 }, { "epoch": 0.22, "learning_rate": 3.6532830959999996e-05, "loss": 0.0664, "step": 70000 }, { "epoch": 0.23, "learning_rate": 3.651735096e-05, "loss": 0.0661, "step": 70500 }, { "epoch": 0.23, "learning_rate": 3.650187096e-05, "loss": 0.0674, "step": 71000 }, { "epoch": 0.23, "learning_rate": 3.648639096e-05, "loss": 0.0649, "step": 71500 }, { "epoch": 0.23, "learning_rate": 3.647091096e-05, "loss": 0.0677, "step": 72000 }, { "epoch": 0.23, "learning_rate": 3.6455430959999995e-05, "loss": 0.0669, "step": 72500 }, { "epoch": 0.23, "learning_rate": 3.643995096e-05, "loss": 0.0646, "step": 73000 }, { "epoch": 0.24, "learning_rate": 3.642447096e-05, "loss": 0.067, "step": 73500 }, { "epoch": 0.24, "learning_rate": 3.640899096e-05, "loss": 0.0669, "step": 74000 }, { "epoch": 0.24, "learning_rate": 3.639351096e-05, "loss": 0.0654, "step": 74500 }, { "epoch": 0.24, "learning_rate": 3.6378030959999995e-05, "loss": 0.067, "step": 75000 }, { "epoch": 0.24, "learning_rate": 3.6362550959999996e-05, "loss": 0.0691, "step": 75500 }, { "epoch": 0.24, "learning_rate": 3.634707096e-05, "loss": 0.066, "step": 76000 }, { "epoch": 0.24, "learning_rate": 3.633159096e-05, "loss": 0.0672, "step": 76500 }, { "epoch": 0.25, "learning_rate": 3.631611096e-05, "loss": 0.0674, "step": 77000 }, { "epoch": 0.25, "learning_rate": 3.6300630959999995e-05, "loss": 0.069, "step": 77500 }, { "epoch": 0.25, "learning_rate": 3.6285150959999996e-05, "loss": 0.068, "step": 78000 }, { "epoch": 0.25, "learning_rate": 3.626967096e-05, "loss": 0.0672, "step": 78500 }, { "epoch": 0.25, "learning_rate": 3.625419096e-05, "loss": 0.0674, "step": 79000 }, { "epoch": 0.25, "learning_rate": 3.623871096e-05, "loss": 0.0669, "step": 79500 }, { "epoch": 0.26, "learning_rate": 3.6223230959999994e-05, "loss": 0.0663, "step": 80000 }, { "epoch": 0.26, "learning_rate": 3.620775096e-05, "loss": 0.0663, "step": 80500 }, { "epoch": 0.26, "learning_rate": 3.6192270960000004e-05, "loss": 0.066, "step": 81000 }, { "epoch": 0.26, "learning_rate": 3.617679096e-05, "loss": 0.066, "step": 81500 }, { "epoch": 0.26, "learning_rate": 3.616131096e-05, "loss": 0.0673, "step": 82000 }, { "epoch": 0.26, "learning_rate": 3.6145830959999994e-05, "loss": 0.0646, "step": 82500 }, { "epoch": 0.27, "learning_rate": 3.613035096e-05, "loss": 0.0662, "step": 83000 }, { "epoch": 0.27, "learning_rate": 3.6114870960000004e-05, "loss": 0.0665, "step": 83500 }, { "epoch": 0.27, "learning_rate": 3.609939096e-05, "loss": 0.0663, "step": 84000 }, { "epoch": 0.27, "learning_rate": 3.608391096e-05, "loss": 0.066, "step": 84500 }, { "epoch": 0.27, "learning_rate": 3.606843096e-05, "loss": 0.0672, "step": 85000 }, { "epoch": 0.27, "learning_rate": 3.605295096e-05, "loss": 0.0658, "step": 85500 }, { "epoch": 0.28, "learning_rate": 3.603747096e-05, "loss": 0.0678, "step": 86000 }, { "epoch": 0.28, "learning_rate": 3.602199096e-05, "loss": 0.0658, "step": 86500 }, { "epoch": 0.28, "learning_rate": 3.600651096e-05, "loss": 0.067, "step": 87000 }, { "epoch": 0.28, "learning_rate": 3.599103096e-05, "loss": 0.0654, "step": 87500 }, { "epoch": 0.28, "learning_rate": 3.597555096e-05, "loss": 0.0661, "step": 88000 }, { "epoch": 0.28, "learning_rate": 3.596007096e-05, "loss": 0.0664, "step": 88500 }, { "epoch": 0.28, "learning_rate": 3.594459096e-05, "loss": 0.0666, "step": 89000 }, { "epoch": 0.29, "learning_rate": 3.592911096e-05, "loss": 0.065, "step": 89500 }, { "epoch": 0.29, "learning_rate": 3.591363096e-05, "loss": 0.066, "step": 90000 }, { "epoch": 0.29, "learning_rate": 3.589815096e-05, "loss": 0.0663, "step": 90500 }, { "epoch": 0.29, "learning_rate": 3.588267096e-05, "loss": 0.064, "step": 91000 }, { "epoch": 0.29, "learning_rate": 3.586719096e-05, "loss": 0.0684, "step": 91500 }, { "epoch": 0.29, "learning_rate": 3.585171096e-05, "loss": 0.0676, "step": 92000 }, { "epoch": 0.3, "learning_rate": 3.583623096e-05, "loss": 0.0681, "step": 92500 }, { "epoch": 0.3, "learning_rate": 3.582075096e-05, "loss": 0.0658, "step": 93000 }, { "epoch": 0.3, "learning_rate": 3.580527096e-05, "loss": 0.0649, "step": 93500 }, { "epoch": 0.3, "learning_rate": 3.578979096e-05, "loss": 0.0678, "step": 94000 }, { "epoch": 0.3, "learning_rate": 3.577431096e-05, "loss": 0.0669, "step": 94500 }, { "epoch": 0.3, "learning_rate": 3.575883096e-05, "loss": 0.0682, "step": 95000 }, { "epoch": 0.31, "learning_rate": 3.574335096e-05, "loss": 0.0655, "step": 95500 }, { "epoch": 0.31, "learning_rate": 3.572787096e-05, "loss": 0.0668, "step": 96000 }, { "epoch": 0.31, "learning_rate": 3.5712390959999996e-05, "loss": 0.0655, "step": 96500 }, { "epoch": 0.31, "learning_rate": 3.569691096e-05, "loss": 0.0671, "step": 97000 }, { "epoch": 0.31, "learning_rate": 3.568143096e-05, "loss": 0.0666, "step": 97500 }, { "epoch": 0.31, "learning_rate": 3.566595096e-05, "loss": 0.0659, "step": 98000 }, { "epoch": 0.32, "learning_rate": 3.565047096e-05, "loss": 0.0647, "step": 98500 }, { "epoch": 0.32, "learning_rate": 3.5634990959999996e-05, "loss": 0.0663, "step": 99000 }, { "epoch": 0.32, "learning_rate": 3.561951096e-05, "loss": 0.0655, "step": 99500 }, { "epoch": 0.32, "learning_rate": 3.560403096e-05, "loss": 0.0656, "step": 100000 }, { "epoch": 0.32, "learning_rate": 3.558855096e-05, "loss": 0.0675, "step": 100500 }, { "epoch": 0.32, "learning_rate": 3.557307096e-05, "loss": 0.0657, "step": 101000 }, { "epoch": 0.32, "learning_rate": 3.5557590959999996e-05, "loss": 0.0669, "step": 101500 }, { "epoch": 0.33, "learning_rate": 3.554211096e-05, "loss": 0.0664, "step": 102000 }, { "epoch": 0.33, "learning_rate": 3.552663096e-05, "loss": 0.0671, "step": 102500 }, { "epoch": 0.33, "learning_rate": 3.551115096e-05, "loss": 0.0652, "step": 103000 }, { "epoch": 0.33, "learning_rate": 3.549567096e-05, "loss": 0.0649, "step": 103500 }, { "epoch": 0.33, "learning_rate": 3.5480190959999996e-05, "loss": 0.0677, "step": 104000 }, { "epoch": 0.33, "learning_rate": 3.546471096e-05, "loss": 0.0651, "step": 104500 }, { "epoch": 0.34, "learning_rate": 3.544923096e-05, "loss": 0.0653, "step": 105000 }, { "epoch": 0.34, "learning_rate": 3.543375096e-05, "loss": 0.0656, "step": 105500 }, { "epoch": 0.34, "learning_rate": 3.541827096e-05, "loss": 0.0663, "step": 106000 }, { "epoch": 0.34, "learning_rate": 3.5402790959999995e-05, "loss": 0.0647, "step": 106500 }, { "epoch": 0.34, "learning_rate": 3.5387310959999997e-05, "loss": 0.0661, "step": 107000 }, { "epoch": 0.34, "learning_rate": 3.537183096e-05, "loss": 0.0661, "step": 107500 }, { "epoch": 0.35, "learning_rate": 3.535635096e-05, "loss": 0.0665, "step": 108000 }, { "epoch": 0.35, "learning_rate": 3.534087096e-05, "loss": 0.0671, "step": 108500 }, { "epoch": 0.35, "learning_rate": 3.5325390959999995e-05, "loss": 0.065, "step": 109000 }, { "epoch": 0.35, "learning_rate": 3.5309910959999996e-05, "loss": 0.067, "step": 109500 }, { "epoch": 0.35, "learning_rate": 3.529443096e-05, "loss": 0.0643, "step": 110000 }, { "epoch": 0.35, "learning_rate": 3.527895096e-05, "loss": 0.0658, "step": 110500 }, { "epoch": 0.36, "learning_rate": 3.526347096e-05, "loss": 0.0654, "step": 111000 }, { "epoch": 0.36, "learning_rate": 3.5247990959999995e-05, "loss": 0.066, "step": 111500 }, { "epoch": 0.36, "learning_rate": 3.5232510959999996e-05, "loss": 0.0654, "step": 112000 }, { "epoch": 0.36, "learning_rate": 3.5217030960000004e-05, "loss": 0.0666, "step": 112500 }, { "epoch": 0.36, "learning_rate": 3.520155096e-05, "loss": 0.0645, "step": 113000 }, { "epoch": 0.36, "learning_rate": 3.518607096e-05, "loss": 0.0646, "step": 113500 }, { "epoch": 0.36, "learning_rate": 3.5170590959999994e-05, "loss": 0.0647, "step": 114000 }, { "epoch": 0.37, "learning_rate": 3.515511096e-05, "loss": 0.0648, "step": 114500 }, { "epoch": 0.37, "learning_rate": 3.5139630960000004e-05, "loss": 0.0661, "step": 115000 }, { "epoch": 0.37, "learning_rate": 3.512415096e-05, "loss": 0.0653, "step": 115500 }, { "epoch": 0.37, "learning_rate": 3.510867096e-05, "loss": 0.0632, "step": 116000 }, { "epoch": 0.37, "learning_rate": 3.509319096e-05, "loss": 0.0643, "step": 116500 }, { "epoch": 0.37, "learning_rate": 3.507771096e-05, "loss": 0.0655, "step": 117000 }, { "epoch": 0.38, "learning_rate": 3.506223096e-05, "loss": 0.0647, "step": 117500 }, { "epoch": 0.38, "learning_rate": 3.504675096e-05, "loss": 0.0641, "step": 118000 }, { "epoch": 0.38, "learning_rate": 3.503127096e-05, "loss": 0.0658, "step": 118500 }, { "epoch": 0.38, "learning_rate": 3.501579096e-05, "loss": 0.0652, "step": 119000 }, { "epoch": 0.38, "learning_rate": 3.500031096e-05, "loss": 0.0653, "step": 119500 }, { "epoch": 0.38, "learning_rate": 3.498483096e-05, "loss": 0.0634, "step": 120000 }, { "epoch": 0.39, "learning_rate": 3.496935096e-05, "loss": 0.0649, "step": 120500 }, { "epoch": 0.39, "learning_rate": 3.495387096e-05, "loss": 0.0647, "step": 121000 }, { "epoch": 0.39, "learning_rate": 3.493839096e-05, "loss": 0.0675, "step": 121500 }, { "epoch": 0.39, "learning_rate": 3.492291096e-05, "loss": 0.0653, "step": 122000 }, { "epoch": 0.39, "learning_rate": 3.490743096e-05, "loss": 0.0636, "step": 122500 }, { "epoch": 0.39, "learning_rate": 3.489195096e-05, "loss": 0.0653, "step": 123000 }, { "epoch": 0.4, "learning_rate": 3.487647096e-05, "loss": 0.0665, "step": 123500 }, { "epoch": 0.4, "learning_rate": 3.486099096e-05, "loss": 0.0641, "step": 124000 }, { "epoch": 0.4, "learning_rate": 3.484551096e-05, "loss": 0.0639, "step": 124500 }, { "epoch": 0.4, "learning_rate": 3.483003096e-05, "loss": 0.0649, "step": 125000 }, { "epoch": 0.4, "learning_rate": 3.481455096e-05, "loss": 0.0655, "step": 125500 }, { "epoch": 0.4, "learning_rate": 3.479907096e-05, "loss": 0.0658, "step": 126000 }, { "epoch": 0.4, "learning_rate": 3.478359096e-05, "loss": 0.0654, "step": 126500 }, { "epoch": 0.41, "learning_rate": 3.476811096e-05, "loss": 0.0656, "step": 127000 }, { "epoch": 0.41, "learning_rate": 3.475263096e-05, "loss": 0.0637, "step": 127500 }, { "epoch": 0.41, "learning_rate": 3.473715096e-05, "loss": 0.0643, "step": 128000 }, { "epoch": 0.41, "learning_rate": 3.472167096e-05, "loss": 0.0658, "step": 128500 }, { "epoch": 0.41, "learning_rate": 3.470619096e-05, "loss": 0.0646, "step": 129000 }, { "epoch": 0.41, "learning_rate": 3.469071096e-05, "loss": 0.0656, "step": 129500 }, { "epoch": 0.42, "learning_rate": 3.467523096e-05, "loss": 0.0637, "step": 130000 }, { "epoch": 0.42, "learning_rate": 3.4659750959999996e-05, "loss": 0.064, "step": 130500 }, { "epoch": 0.42, "learning_rate": 3.464427096e-05, "loss": 0.0619, "step": 131000 }, { "epoch": 0.42, "learning_rate": 3.462879096e-05, "loss": 0.065, "step": 131500 }, { "epoch": 0.42, "learning_rate": 3.461331096e-05, "loss": 0.0657, "step": 132000 }, { "epoch": 0.42, "learning_rate": 3.459783096e-05, "loss": 0.0639, "step": 132500 }, { "epoch": 0.43, "learning_rate": 3.4582350959999996e-05, "loss": 0.0642, "step": 133000 }, { "epoch": 0.43, "learning_rate": 3.456687096e-05, "loss": 0.0649, "step": 133500 }, { "epoch": 0.43, "learning_rate": 3.455139096e-05, "loss": 0.065, "step": 134000 }, { "epoch": 0.43, "learning_rate": 3.453591096e-05, "loss": 0.0652, "step": 134500 }, { "epoch": 0.43, "learning_rate": 3.452043096e-05, "loss": 0.062, "step": 135000 }, { "epoch": 0.43, "learning_rate": 3.4504950959999996e-05, "loss": 0.0641, "step": 135500 }, { "epoch": 0.44, "learning_rate": 3.448947096e-05, "loss": 0.0648, "step": 136000 }, { "epoch": 0.44, "learning_rate": 3.447399096e-05, "loss": 0.0623, "step": 136500 }, { "epoch": 0.44, "learning_rate": 3.445851096e-05, "loss": 0.0649, "step": 137000 }, { "epoch": 0.44, "learning_rate": 3.444303096e-05, "loss": 0.0639, "step": 137500 }, { "epoch": 0.44, "learning_rate": 3.4427550959999995e-05, "loss": 0.0618, "step": 138000 }, { "epoch": 0.44, "learning_rate": 3.441207096e-05, "loss": 0.0639, "step": 138500 }, { "epoch": 0.44, "learning_rate": 3.439659096e-05, "loss": 0.0657, "step": 139000 }, { "epoch": 0.45, "learning_rate": 3.438111096e-05, "loss": 0.0635, "step": 139500 }, { "epoch": 0.45, "learning_rate": 3.436563096e-05, "loss": 0.0651, "step": 140000 }, { "epoch": 0.45, "learning_rate": 3.4350150959999995e-05, "loss": 0.0644, "step": 140500 }, { "epoch": 0.45, "learning_rate": 3.4334670959999996e-05, "loss": 0.064, "step": 141000 }, { "epoch": 0.45, "learning_rate": 3.431919096e-05, "loss": 0.0641, "step": 141500 }, { "epoch": 0.45, "learning_rate": 3.430371096e-05, "loss": 0.0633, "step": 142000 }, { "epoch": 0.46, "learning_rate": 3.428823096e-05, "loss": 0.0648, "step": 142500 }, { "epoch": 0.46, "learning_rate": 3.4272750959999995e-05, "loss": 0.064, "step": 143000 }, { "epoch": 0.46, "learning_rate": 3.4257270959999996e-05, "loss": 0.0634, "step": 143500 }, { "epoch": 0.46, "learning_rate": 3.4241790960000004e-05, "loss": 0.062, "step": 144000 }, { "epoch": 0.46, "learning_rate": 3.422631096e-05, "loss": 0.0627, "step": 144500 }, { "epoch": 0.46, "learning_rate": 3.421083096e-05, "loss": 0.0635, "step": 145000 }, { "epoch": 0.47, "learning_rate": 3.4195350959999994e-05, "loss": 0.0646, "step": 145500 }, { "epoch": 0.47, "learning_rate": 3.417987096e-05, "loss": 0.0636, "step": 146000 }, { "epoch": 0.47, "learning_rate": 3.4164390960000004e-05, "loss": 0.0623, "step": 146500 }, { "epoch": 0.47, "learning_rate": 3.414891096e-05, "loss": 0.0649, "step": 147000 }, { "epoch": 0.47, "learning_rate": 3.413343096e-05, "loss": 0.0651, "step": 147500 }, { "epoch": 0.47, "learning_rate": 3.4117950959999994e-05, "loss": 0.0626, "step": 148000 }, { "epoch": 0.48, "learning_rate": 3.410247096e-05, "loss": 0.0642, "step": 148500 }, { "epoch": 0.48, "learning_rate": 3.4086990960000004e-05, "loss": 0.0636, "step": 149000 }, { "epoch": 0.48, "learning_rate": 3.407151096e-05, "loss": 0.0616, "step": 149500 }, { "epoch": 0.48, "learning_rate": 3.405603096e-05, "loss": 0.0641, "step": 150000 }, { "epoch": 0.48, "learning_rate": 3.404055096e-05, "loss": 0.0626, "step": 150500 }, { "epoch": 0.48, "learning_rate": 3.402507096e-05, "loss": 0.061, "step": 151000 }, { "epoch": 0.48, "learning_rate": 3.400959096e-05, "loss": 0.063, "step": 151500 }, { "epoch": 0.49, "learning_rate": 3.399411096e-05, "loss": 0.065, "step": 152000 }, { "epoch": 0.49, "learning_rate": 3.397863096e-05, "loss": 0.0655, "step": 152500 }, { "epoch": 0.49, "learning_rate": 3.396315096e-05, "loss": 0.0632, "step": 153000 }, { "epoch": 0.49, "learning_rate": 3.394767096e-05, "loss": 0.0626, "step": 153500 }, { "epoch": 0.49, "learning_rate": 3.393219096e-05, "loss": 0.064, "step": 154000 }, { "epoch": 0.49, "learning_rate": 3.391671096e-05, "loss": 0.0641, "step": 154500 }, { "epoch": 0.5, "learning_rate": 3.390123096e-05, "loss": 0.063, "step": 155000 }, { "epoch": 0.5, "learning_rate": 3.388575096e-05, "loss": 0.0632, "step": 155500 }, { "epoch": 0.5, "learning_rate": 3.387027096e-05, "loss": 0.0629, "step": 156000 }, { "epoch": 0.5, "learning_rate": 3.385479096e-05, "loss": 0.0637, "step": 156500 }, { "epoch": 0.5, "learning_rate": 3.383931096e-05, "loss": 0.0644, "step": 157000 }, { "epoch": 0.5, "learning_rate": 3.382383096e-05, "loss": 0.0642, "step": 157500 }, { "epoch": 0.51, "learning_rate": 3.380835096e-05, "loss": 0.0621, "step": 158000 }, { "epoch": 0.51, "learning_rate": 3.379287096e-05, "loss": 0.0624, "step": 158500 }, { "epoch": 0.51, "learning_rate": 3.377739096e-05, "loss": 0.0658, "step": 159000 }, { "epoch": 0.51, "learning_rate": 3.376191096e-05, "loss": 0.0651, "step": 159500 }, { "epoch": 0.51, "learning_rate": 3.374643096e-05, "loss": 0.0624, "step": 160000 }, { "epoch": 0.51, "learning_rate": 3.373095096e-05, "loss": 0.0629, "step": 160500 }, { "epoch": 0.52, "learning_rate": 3.371547096e-05, "loss": 0.0635, "step": 161000 }, { "epoch": 0.52, "learning_rate": 3.369999096e-05, "loss": 0.063, "step": 161500 }, { "epoch": 0.52, "learning_rate": 3.3684510959999997e-05, "loss": 0.0629, "step": 162000 }, { "epoch": 0.52, "learning_rate": 3.366903096e-05, "loss": 0.0629, "step": 162500 }, { "epoch": 0.52, "learning_rate": 3.365355096e-05, "loss": 0.0649, "step": 163000 }, { "epoch": 0.52, "learning_rate": 3.363807096e-05, "loss": 0.0608, "step": 163500 }, { "epoch": 0.52, "learning_rate": 3.362259096e-05, "loss": 0.0624, "step": 164000 }, { "epoch": 0.53, "learning_rate": 3.3607110959999996e-05, "loss": 0.065, "step": 164500 }, { "epoch": 0.53, "learning_rate": 3.359163096e-05, "loss": 0.0636, "step": 165000 }, { "epoch": 0.53, "learning_rate": 3.357615096e-05, "loss": 0.0629, "step": 165500 }, { "epoch": 0.53, "learning_rate": 3.356067096e-05, "loss": 0.0639, "step": 166000 }, { "epoch": 0.53, "learning_rate": 3.354519096e-05, "loss": 0.0639, "step": 166500 }, { "epoch": 0.53, "learning_rate": 3.3529710959999996e-05, "loss": 0.0638, "step": 167000 }, { "epoch": 0.54, "learning_rate": 3.351423096e-05, "loss": 0.0628, "step": 167500 }, { "epoch": 0.54, "learning_rate": 3.349875096e-05, "loss": 0.0645, "step": 168000 }, { "epoch": 0.54, "learning_rate": 3.348327096e-05, "loss": 0.0626, "step": 168500 }, { "epoch": 0.54, "learning_rate": 3.346779096e-05, "loss": 0.0638, "step": 169000 }, { "epoch": 0.54, "learning_rate": 3.3452310959999996e-05, "loss": 0.0632, "step": 169500 }, { "epoch": 0.54, "learning_rate": 3.343683096e-05, "loss": 0.0631, "step": 170000 }, { "epoch": 0.55, "learning_rate": 3.342135096e-05, "loss": 0.0637, "step": 170500 }, { "epoch": 0.55, "learning_rate": 3.340587096e-05, "loss": 0.0626, "step": 171000 }, { "epoch": 0.55, "learning_rate": 3.339039096e-05, "loss": 0.0633, "step": 171500 }, { "epoch": 0.55, "learning_rate": 3.3374910959999995e-05, "loss": 0.0622, "step": 172000 }, { "epoch": 0.55, "learning_rate": 3.3359430959999997e-05, "loss": 0.062, "step": 172500 }, { "epoch": 0.55, "learning_rate": 3.334395096e-05, "loss": 0.063, "step": 173000 }, { "epoch": 0.56, "learning_rate": 3.332847096e-05, "loss": 0.0631, "step": 173500 }, { "epoch": 0.56, "learning_rate": 3.331299096e-05, "loss": 0.0637, "step": 174000 }, { "epoch": 0.56, "learning_rate": 3.3297510959999995e-05, "loss": 0.0633, "step": 174500 }, { "epoch": 0.56, "learning_rate": 3.3282030959999996e-05, "loss": 0.0628, "step": 175000 }, { "epoch": 0.56, "learning_rate": 3.3266550960000004e-05, "loss": 0.0633, "step": 175500 }, { "epoch": 0.56, "learning_rate": 3.325107096e-05, "loss": 0.0638, "step": 176000 }, { "epoch": 0.56, "learning_rate": 3.323559096e-05, "loss": 0.0635, "step": 176500 }, { "epoch": 0.57, "learning_rate": 3.3220110959999995e-05, "loss": 0.064, "step": 177000 }, { "epoch": 0.57, "learning_rate": 3.3204630959999996e-05, "loss": 0.0628, "step": 177500 }, { "epoch": 0.57, "learning_rate": 3.3189150960000004e-05, "loss": 0.0621, "step": 178000 }, { "epoch": 0.57, "learning_rate": 3.317367096e-05, "loss": 0.0614, "step": 178500 }, { "epoch": 0.57, "learning_rate": 3.315819096e-05, "loss": 0.0635, "step": 179000 }, { "epoch": 0.57, "learning_rate": 3.3142710959999994e-05, "loss": 0.0642, "step": 179500 }, { "epoch": 0.58, "learning_rate": 3.312723096e-05, "loss": 0.0617, "step": 180000 }, { "epoch": 0.58, "learning_rate": 3.3111750960000004e-05, "loss": 0.0622, "step": 180500 }, { "epoch": 0.58, "learning_rate": 3.309627096e-05, "loss": 0.0634, "step": 181000 }, { "epoch": 0.58, "learning_rate": 3.308079096e-05, "loss": 0.0614, "step": 181500 }, { "epoch": 0.58, "learning_rate": 3.3065310959999994e-05, "loss": 0.0627, "step": 182000 }, { "epoch": 0.58, "learning_rate": 3.304983096e-05, "loss": 0.0626, "step": 182500 }, { "epoch": 0.59, "learning_rate": 3.3034350960000003e-05, "loss": 0.0623, "step": 183000 }, { "epoch": 0.59, "learning_rate": 3.301887096e-05, "loss": 0.0636, "step": 183500 }, { "epoch": 0.59, "learning_rate": 3.300339096e-05, "loss": 0.0612, "step": 184000 }, { "epoch": 0.59, "learning_rate": 3.298791096e-05, "loss": 0.0623, "step": 184500 }, { "epoch": 0.59, "learning_rate": 3.297243096e-05, "loss": 0.0629, "step": 185000 }, { "epoch": 0.59, "learning_rate": 3.295695096e-05, "loss": 0.062, "step": 185500 }, { "epoch": 0.6, "learning_rate": 3.294147096e-05, "loss": 0.0637, "step": 186000 }, { "epoch": 0.6, "learning_rate": 3.292599096e-05, "loss": 0.0629, "step": 186500 }, { "epoch": 0.6, "learning_rate": 3.291051096e-05, "loss": 0.0617, "step": 187000 }, { "epoch": 0.6, "learning_rate": 3.289503096e-05, "loss": 0.0628, "step": 187500 }, { "epoch": 0.6, "learning_rate": 3.287955096e-05, "loss": 0.0617, "step": 188000 }, { "epoch": 0.6, "learning_rate": 3.286407096e-05, "loss": 0.0637, "step": 188500 }, { "epoch": 0.6, "learning_rate": 3.284859096e-05, "loss": 0.0619, "step": 189000 }, { "epoch": 0.61, "learning_rate": 3.283311096e-05, "loss": 0.0628, "step": 189500 }, { "epoch": 0.61, "learning_rate": 3.281763096e-05, "loss": 0.0638, "step": 190000 }, { "epoch": 0.61, "learning_rate": 3.280215096e-05, "loss": 0.0628, "step": 190500 }, { "epoch": 0.61, "learning_rate": 3.278667096e-05, "loss": 0.0628, "step": 191000 }, { "epoch": 0.61, "learning_rate": 3.277119096e-05, "loss": 0.0616, "step": 191500 }, { "epoch": 0.61, "learning_rate": 3.275571096e-05, "loss": 0.0623, "step": 192000 }, { "epoch": 0.62, "learning_rate": 3.274023096e-05, "loss": 0.0621, "step": 192500 }, { "epoch": 0.62, "learning_rate": 3.272475096e-05, "loss": 0.0619, "step": 193000 }, { "epoch": 0.62, "learning_rate": 3.270927096e-05, "loss": 0.0623, "step": 193500 }, { "epoch": 0.62, "learning_rate": 3.269379096e-05, "loss": 0.063, "step": 194000 }, { "epoch": 0.62, "learning_rate": 3.267831096e-05, "loss": 0.0629, "step": 194500 }, { "epoch": 0.62, "learning_rate": 3.266283096e-05, "loss": 0.0599, "step": 195000 }, { "epoch": 0.63, "learning_rate": 3.264735096e-05, "loss": 0.062, "step": 195500 }, { "epoch": 0.63, "learning_rate": 3.2631870959999996e-05, "loss": 0.0633, "step": 196000 }, { "epoch": 0.63, "learning_rate": 3.261639096e-05, "loss": 0.0606, "step": 196500 }, { "epoch": 0.63, "learning_rate": 3.260091096e-05, "loss": 0.0633, "step": 197000 }, { "epoch": 0.63, "learning_rate": 3.258543096e-05, "loss": 0.0622, "step": 197500 }, { "epoch": 0.63, "learning_rate": 3.256995096e-05, "loss": 0.061, "step": 198000 }, { "epoch": 0.64, "learning_rate": 3.2554470959999996e-05, "loss": 0.0631, "step": 198500 }, { "epoch": 0.64, "learning_rate": 3.253899096e-05, "loss": 0.0631, "step": 199000 }, { "epoch": 0.64, "learning_rate": 3.252351096e-05, "loss": 0.0632, "step": 199500 }, { "epoch": 0.64, "learning_rate": 3.250803096e-05, "loss": 0.0624, "step": 200000 }, { "epoch": 0.64, "learning_rate": 3.249255096e-05, "loss": 0.0607, "step": 200500 }, { "epoch": 0.64, "learning_rate": 3.2477070959999996e-05, "loss": 0.0619, "step": 201000 }, { "epoch": 0.64, "learning_rate": 3.246159096e-05, "loss": 0.0614, "step": 201500 }, { "epoch": 0.65, "learning_rate": 3.244611096e-05, "loss": 0.0606, "step": 202000 }, { "epoch": 0.65, "learning_rate": 3.243063096e-05, "loss": 0.0617, "step": 202500 }, { "epoch": 0.65, "learning_rate": 3.241515096e-05, "loss": 0.0661, "step": 203000 }, { "epoch": 0.65, "learning_rate": 3.2399670959999995e-05, "loss": 0.0617, "step": 203500 }, { "epoch": 0.65, "learning_rate": 3.238419096e-05, "loss": 0.0617, "step": 204000 }, { "epoch": 0.65, "learning_rate": 3.236871096e-05, "loss": 0.0623, "step": 204500 }, { "epoch": 0.66, "learning_rate": 3.235323096e-05, "loss": 0.061, "step": 205000 }, { "epoch": 0.66, "learning_rate": 3.233775096e-05, "loss": 0.0615, "step": 205500 }, { "epoch": 0.66, "learning_rate": 3.2322270959999995e-05, "loss": 0.0615, "step": 206000 }, { "epoch": 0.66, "learning_rate": 3.2306790959999996e-05, "loss": 0.063, "step": 206500 }, { "epoch": 0.66, "learning_rate": 3.229131096e-05, "loss": 0.0615, "step": 207000 }, { "epoch": 0.66, "learning_rate": 3.227583096e-05, "loss": 0.0626, "step": 207500 }, { "epoch": 0.67, "learning_rate": 3.226035096e-05, "loss": 0.0615, "step": 208000 }, { "epoch": 0.67, "learning_rate": 3.2244870959999995e-05, "loss": 0.0632, "step": 208500 }, { "epoch": 0.67, "learning_rate": 3.2229390959999996e-05, "loss": 0.0633, "step": 209000 }, { "epoch": 0.67, "learning_rate": 3.2213910960000004e-05, "loss": 0.062, "step": 209500 }, { "epoch": 0.67, "learning_rate": 3.219843096e-05, "loss": 0.0626, "step": 210000 }, { "epoch": 0.67, "learning_rate": 3.218295096e-05, "loss": 0.0608, "step": 210500 }, { "epoch": 0.68, "learning_rate": 3.2167470959999995e-05, "loss": 0.06, "step": 211000 }, { "epoch": 0.68, "learning_rate": 3.215199096e-05, "loss": 0.0619, "step": 211500 }, { "epoch": 0.68, "learning_rate": 3.2136510960000004e-05, "loss": 0.0598, "step": 212000 }, { "epoch": 0.68, "learning_rate": 3.212103096e-05, "loss": 0.0614, "step": 212500 }, { "epoch": 0.68, "learning_rate": 3.210555096e-05, "loss": 0.0627, "step": 213000 }, { "epoch": 0.68, "learning_rate": 3.2090070959999994e-05, "loss": 0.0611, "step": 213500 }, { "epoch": 0.68, "learning_rate": 3.207459096e-05, "loss": 0.0613, "step": 214000 }, { "epoch": 0.69, "learning_rate": 3.2059110960000004e-05, "loss": 0.0597, "step": 214500 }, { "epoch": 0.69, "learning_rate": 3.204363096e-05, "loss": 0.0618, "step": 215000 }, { "epoch": 0.69, "learning_rate": 3.202815096e-05, "loss": 0.0632, "step": 215500 }, { "epoch": 0.69, "learning_rate": 3.201267096e-05, "loss": 0.0621, "step": 216000 }, { "epoch": 0.69, "learning_rate": 3.199719096e-05, "loss": 0.0627, "step": 216500 }, { "epoch": 0.69, "learning_rate": 3.198171096e-05, "loss": 0.0581, "step": 217000 }, { "epoch": 0.7, "learning_rate": 3.196623096e-05, "loss": 0.0609, "step": 217500 }, { "epoch": 0.7, "learning_rate": 3.195075096e-05, "loss": 0.0601, "step": 218000 }, { "epoch": 0.7, "learning_rate": 3.193527096e-05, "loss": 0.0614, "step": 218500 }, { "epoch": 0.7, "learning_rate": 3.191979096e-05, "loss": 0.0621, "step": 219000 }, { "epoch": 0.7, "learning_rate": 3.190431096e-05, "loss": 0.0626, "step": 219500 }, { "epoch": 0.7, "learning_rate": 3.188883096e-05, "loss": 0.0611, "step": 220000 }, { "epoch": 0.71, "learning_rate": 3.187335096e-05, "loss": 0.0614, "step": 220500 }, { "epoch": 0.71, "learning_rate": 3.185787096e-05, "loss": 0.061, "step": 221000 }, { "epoch": 0.71, "learning_rate": 3.184239096e-05, "loss": 0.0616, "step": 221500 }, { "epoch": 0.71, "learning_rate": 3.182691096e-05, "loss": 0.062, "step": 222000 }, { "epoch": 0.71, "learning_rate": 3.181143096e-05, "loss": 0.0623, "step": 222500 }, { "epoch": 0.71, "learning_rate": 3.179595096e-05, "loss": 0.0603, "step": 223000 }, { "epoch": 0.72, "learning_rate": 3.178047096e-05, "loss": 0.0623, "step": 223500 }, { "epoch": 0.72, "learning_rate": 3.176499096e-05, "loss": 0.0605, "step": 224000 }, { "epoch": 0.72, "learning_rate": 3.174951096e-05, "loss": 0.0614, "step": 224500 }, { "epoch": 0.72, "learning_rate": 3.173403096e-05, "loss": 0.0599, "step": 225000 }, { "epoch": 0.72, "learning_rate": 3.171855096e-05, "loss": 0.0635, "step": 225500 }, { "epoch": 0.72, "learning_rate": 3.170307096e-05, "loss": 0.0593, "step": 226000 }, { "epoch": 0.72, "learning_rate": 3.168759096e-05, "loss": 0.0601, "step": 226500 }, { "epoch": 0.73, "learning_rate": 3.167211096e-05, "loss": 0.0601, "step": 227000 }, { "epoch": 0.73, "learning_rate": 3.1656630959999997e-05, "loss": 0.061, "step": 227500 }, { "epoch": 0.73, "learning_rate": 3.164115096e-05, "loss": 0.0628, "step": 228000 }, { "epoch": 0.73, "learning_rate": 3.162567096e-05, "loss": 0.0618, "step": 228500 }, { "epoch": 0.73, "learning_rate": 3.161019096e-05, "loss": 0.0616, "step": 229000 }, { "epoch": 0.73, "learning_rate": 3.159471096e-05, "loss": 0.0607, "step": 229500 }, { "epoch": 0.74, "learning_rate": 3.1579230959999996e-05, "loss": 0.0605, "step": 230000 }, { "epoch": 0.74, "learning_rate": 3.156375096e-05, "loss": 0.0617, "step": 230500 }, { "epoch": 0.74, "learning_rate": 3.154827096e-05, "loss": 0.0626, "step": 231000 }, { "epoch": 0.74, "learning_rate": 3.153279096e-05, "loss": 0.0603, "step": 231500 }, { "epoch": 0.74, "learning_rate": 3.151731096e-05, "loss": 0.061, "step": 232000 }, { "epoch": 0.74, "learning_rate": 3.1501830959999996e-05, "loss": 0.0616, "step": 232500 }, { "epoch": 0.75, "learning_rate": 3.148635096e-05, "loss": 0.0618, "step": 233000 }, { "epoch": 0.75, "learning_rate": 3.147087096e-05, "loss": 0.0603, "step": 233500 }, { "epoch": 0.75, "learning_rate": 3.145539096e-05, "loss": 0.0595, "step": 234000 }, { "epoch": 0.75, "learning_rate": 3.143991096e-05, "loss": 0.0608, "step": 234500 }, { "epoch": 0.75, "learning_rate": 3.1424430959999996e-05, "loss": 0.0626, "step": 235000 }, { "epoch": 0.75, "learning_rate": 3.140895096e-05, "loss": 0.0625, "step": 235500 }, { "epoch": 0.76, "learning_rate": 3.139347096e-05, "loss": 0.0617, "step": 236000 }, { "epoch": 0.76, "learning_rate": 3.137799096e-05, "loss": 0.0609, "step": 236500 }, { "epoch": 0.76, "learning_rate": 3.136251096e-05, "loss": 0.062, "step": 237000 }, { "epoch": 0.76, "learning_rate": 3.1347030959999995e-05, "loss": 0.0604, "step": 237500 }, { "epoch": 0.76, "learning_rate": 3.1331550959999997e-05, "loss": 0.0597, "step": 238000 }, { "epoch": 0.76, "learning_rate": 3.131607096e-05, "loss": 0.0608, "step": 238500 }, { "epoch": 0.76, "learning_rate": 3.130059096e-05, "loss": 0.0627, "step": 239000 }, { "epoch": 0.77, "learning_rate": 3.128511096e-05, "loss": 0.0609, "step": 239500 }, { "epoch": 0.77, "learning_rate": 3.1269630959999995e-05, "loss": 0.061, "step": 240000 }, { "epoch": 0.77, "learning_rate": 3.1254150959999996e-05, "loss": 0.062, "step": 240500 }, { "epoch": 0.77, "learning_rate": 3.1238670960000004e-05, "loss": 0.0612, "step": 241000 }, { "epoch": 0.77, "learning_rate": 3.122319096e-05, "loss": 0.0621, "step": 241500 }, { "epoch": 0.77, "learning_rate": 3.120771096e-05, "loss": 0.0604, "step": 242000 }, { "epoch": 0.78, "learning_rate": 3.1192230959999995e-05, "loss": 0.0614, "step": 242500 }, { "epoch": 0.78, "learning_rate": 3.1176750959999996e-05, "loss": 0.0598, "step": 243000 }, { "epoch": 0.78, "learning_rate": 3.1161270960000004e-05, "loss": 0.0623, "step": 243500 }, { "epoch": 0.78, "learning_rate": 3.114579096e-05, "loss": 0.0626, "step": 244000 }, { "epoch": 0.78, "learning_rate": 3.113031096e-05, "loss": 0.0592, "step": 244500 }, { "epoch": 0.78, "learning_rate": 3.1114830959999994e-05, "loss": 0.0599, "step": 245000 }, { "epoch": 0.79, "learning_rate": 3.109935096e-05, "loss": 0.0612, "step": 245500 }, { "epoch": 0.79, "learning_rate": 3.1083870960000004e-05, "loss": 0.0605, "step": 246000 }, { "epoch": 0.79, "learning_rate": 3.106839096e-05, "loss": 0.0617, "step": 246500 }, { "epoch": 0.79, "learning_rate": 3.105291096e-05, "loss": 0.0597, "step": 247000 }, { "epoch": 0.79, "learning_rate": 3.1037430959999994e-05, "loss": 0.0605, "step": 247500 }, { "epoch": 0.79, "learning_rate": 3.102195096e-05, "loss": 0.0595, "step": 248000 }, { "epoch": 0.8, "learning_rate": 3.1006470960000003e-05, "loss": 0.0598, "step": 248500 }, { "epoch": 0.8, "learning_rate": 3.099099096e-05, "loss": 0.06, "step": 249000 }, { "epoch": 0.8, "learning_rate": 3.097551096e-05, "loss": 0.0615, "step": 249500 }, { "epoch": 0.8, "learning_rate": 3.096003096e-05, "loss": 0.0613, "step": 250000 }, { "epoch": 0.8, "learning_rate": 3.094455096e-05, "loss": 0.0587, "step": 250500 }, { "epoch": 0.8, "learning_rate": 3.092907096e-05, "loss": 0.0604, "step": 251000 }, { "epoch": 0.8, "learning_rate": 3.091359096e-05, "loss": 0.0608, "step": 251500 }, { "epoch": 0.81, "learning_rate": 3.089811096e-05, "loss": 0.0598, "step": 252000 }, { "epoch": 0.81, "learning_rate": 3.088263096e-05, "loss": 0.0602, "step": 252500 }, { "epoch": 0.81, "learning_rate": 3.086715096e-05, "loss": 0.0586, "step": 253000 }, { "epoch": 0.81, "learning_rate": 3.085167096e-05, "loss": 0.0598, "step": 253500 }, { "epoch": 0.81, "learning_rate": 3.083619096e-05, "loss": 0.0618, "step": 254000 }, { "epoch": 0.81, "learning_rate": 3.082071096e-05, "loss": 0.0604, "step": 254500 }, { "epoch": 0.82, "learning_rate": 3.080523096e-05, "loss": 0.0604, "step": 255000 }, { "epoch": 0.82, "learning_rate": 3.078975096e-05, "loss": 0.0601, "step": 255500 }, { "epoch": 0.82, "learning_rate": 3.077427096e-05, "loss": 0.0593, "step": 256000 }, { "epoch": 0.82, "learning_rate": 3.075879096e-05, "loss": 0.0593, "step": 256500 }, { "epoch": 0.82, "learning_rate": 3.074331096e-05, "loss": 0.0592, "step": 257000 }, { "epoch": 0.82, "learning_rate": 3.072783096e-05, "loss": 0.0594, "step": 257500 }, { "epoch": 0.83, "learning_rate": 3.071235096e-05, "loss": 0.0623, "step": 258000 }, { "epoch": 0.83, "learning_rate": 3.069687096e-05, "loss": 0.0613, "step": 258500 }, { "epoch": 0.83, "learning_rate": 3.068139096e-05, "loss": 0.0605, "step": 259000 }, { "epoch": 0.83, "learning_rate": 3.066591096e-05, "loss": 0.0624, "step": 259500 }, { "epoch": 0.83, "learning_rate": 3.065043096e-05, "loss": 0.0595, "step": 260000 }, { "epoch": 0.83, "learning_rate": 3.063495096e-05, "loss": 0.0603, "step": 260500 }, { "epoch": 0.84, "learning_rate": 3.061947096e-05, "loss": 0.0594, "step": 261000 }, { "epoch": 0.84, "learning_rate": 3.0603990959999996e-05, "loss": 0.0628, "step": 261500 }, { "epoch": 0.84, "learning_rate": 3.058851096e-05, "loss": 0.0592, "step": 262000 }, { "epoch": 0.84, "learning_rate": 3.057303096e-05, "loss": 0.0596, "step": 262500 }, { "epoch": 0.84, "learning_rate": 3.055755096e-05, "loss": 0.0589, "step": 263000 }, { "epoch": 0.84, "learning_rate": 3.054207096e-05, "loss": 0.0595, "step": 263500 }, { "epoch": 0.84, "learning_rate": 3.0526590959999996e-05, "loss": 0.062, "step": 264000 }, { "epoch": 0.85, "learning_rate": 3.051111096e-05, "loss": 0.06, "step": 264500 }, { "epoch": 0.85, "learning_rate": 3.0495630959999995e-05, "loss": 0.0593, "step": 265000 }, { "epoch": 0.85, "learning_rate": 3.048015096e-05, "loss": 0.0586, "step": 265500 }, { "epoch": 0.85, "learning_rate": 3.046467096e-05, "loss": 0.0604, "step": 266000 }, { "epoch": 0.85, "learning_rate": 3.044919096e-05, "loss": 0.0595, "step": 266500 }, { "epoch": 0.85, "learning_rate": 3.043371096e-05, "loss": 0.0606, "step": 267000 }, { "epoch": 0.86, "learning_rate": 3.0418230960000002e-05, "loss": 0.0606, "step": 267500 }, { "epoch": 0.86, "learning_rate": 3.040275096e-05, "loss": 0.0614, "step": 268000 }, { "epoch": 0.86, "learning_rate": 3.038727096e-05, "loss": 0.0584, "step": 268500 }, { "epoch": 0.86, "learning_rate": 3.037179096e-05, "loss": 0.0606, "step": 269000 }, { "epoch": 0.86, "learning_rate": 3.035631096e-05, "loss": 0.0585, "step": 269500 }, { "epoch": 0.86, "learning_rate": 3.034083096e-05, "loss": 0.0589, "step": 270000 }, { "epoch": 0.87, "learning_rate": 3.032535096e-05, "loss": 0.0581, "step": 270500 }, { "epoch": 0.87, "learning_rate": 3.030987096e-05, "loss": 0.0582, "step": 271000 }, { "epoch": 0.87, "learning_rate": 3.029439096e-05, "loss": 0.0606, "step": 271500 }, { "epoch": 0.87, "learning_rate": 3.027891096e-05, "loss": 0.0586, "step": 272000 }, { "epoch": 0.87, "learning_rate": 3.026343096e-05, "loss": 0.0594, "step": 272500 }, { "epoch": 0.87, "learning_rate": 3.024795096e-05, "loss": 0.0586, "step": 273000 }, { "epoch": 0.88, "learning_rate": 3.023247096e-05, "loss": 0.0589, "step": 273500 }, { "epoch": 0.88, "learning_rate": 3.0216990959999998e-05, "loss": 0.0589, "step": 274000 }, { "epoch": 0.88, "learning_rate": 3.020151096e-05, "loss": 0.0595, "step": 274500 }, { "epoch": 0.88, "learning_rate": 3.018603096e-05, "loss": 0.0599, "step": 275000 }, { "epoch": 0.88, "learning_rate": 3.017055096e-05, "loss": 0.0596, "step": 275500 }, { "epoch": 0.88, "learning_rate": 3.015507096e-05, "loss": 0.0591, "step": 276000 }, { "epoch": 0.88, "learning_rate": 3.0139590959999998e-05, "loss": 0.0587, "step": 276500 }, { "epoch": 0.89, "learning_rate": 3.012411096e-05, "loss": 0.0595, "step": 277000 }, { "epoch": 0.89, "learning_rate": 3.010863096e-05, "loss": 0.0597, "step": 277500 }, { "epoch": 0.89, "learning_rate": 3.009315096e-05, "loss": 0.0595, "step": 278000 }, { "epoch": 0.89, "learning_rate": 3.007767096e-05, "loss": 0.0585, "step": 278500 }, { "epoch": 0.89, "learning_rate": 3.0062190959999998e-05, "loss": 0.0597, "step": 279000 }, { "epoch": 0.89, "learning_rate": 3.004671096e-05, "loss": 0.0596, "step": 279500 }, { "epoch": 0.9, "learning_rate": 3.003123096e-05, "loss": 0.0592, "step": 280000 } ], "max_steps": 1250000, "num_train_epochs": 4, "total_flos": 4.30189365889536e+17, "trial_name": null, "trial_params": null }