{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8457359110620398, "eval_steps": 100000000, "global_step": 91000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 5e-05, "loss": 7.7296, "step": 500 }, { "epoch": 0.0, "learning_rate": 4.9999535305620204e-05, "loss": 6.5786, "step": 1000 }, { "epoch": 0.0, "learning_rate": 4.99990706112404e-05, "loss": 6.2317, "step": 1500 }, { "epoch": 0.0, "learning_rate": 4.99986059168606e-05, "loss": 6.0053, "step": 2000 }, { "epoch": 0.0, "learning_rate": 4.99981412224808e-05, "loss": 5.8615, "step": 2500 }, { "epoch": 0.0, "learning_rate": 4.9997676528101e-05, "loss": 5.6794, "step": 3000 }, { "epoch": 0.0, "learning_rate": 4.99972118337212e-05, "loss": 5.5481, "step": 3500 }, { "epoch": 0.0, "learning_rate": 4.9996747139341396e-05, "loss": 5.4375, "step": 4000 }, { "epoch": 0.0, "learning_rate": 4.99962824449616e-05, "loss": 5.3391, "step": 4500 }, { "epoch": 0.0, "learning_rate": 4.99958177505818e-05, "loss": 5.2598, "step": 5000 }, { "epoch": 0.01, "learning_rate": 4.9995353056202e-05, "loss": 5.17, "step": 5500 }, { "epoch": 0.01, "learning_rate": 4.9994888361822204e-05, "loss": 5.0886, "step": 6000 }, { "epoch": 0.01, "learning_rate": 4.99944236674424e-05, "loss": 5.0128, "step": 6500 }, { "epoch": 0.01, "learning_rate": 4.99939589730626e-05, "loss": 4.9504, "step": 7000 }, { "epoch": 0.01, "learning_rate": 4.99934942786828e-05, "loss": 4.8755, "step": 7500 }, { "epoch": 0.01, "learning_rate": 4.9993029584303e-05, "loss": 4.8402, "step": 8000 }, { "epoch": 0.01, "learning_rate": 4.99925648899232e-05, "loss": 4.7674, "step": 8500 }, { "epoch": 0.01, "learning_rate": 4.9992100195543396e-05, "loss": 4.7254, "step": 9000 }, { "epoch": 0.01, "learning_rate": 4.99916355011636e-05, "loss": 4.6749, "step": 9500 }, { "epoch": 0.01, "learning_rate": 4.99911708067838e-05, "loss": 4.6135, "step": 10000 }, { "epoch": 0.01, "learning_rate": 4.9990706112403995e-05, "loss": 4.5239, "step": 10500 }, { "epoch": 0.01, "learning_rate": 4.99902414180242e-05, "loss": 4.4729, "step": 11000 }, { "epoch": 0.01, "learning_rate": 4.998977672364439e-05, "loss": 4.4434, "step": 11500 }, { "epoch": 0.01, "learning_rate": 4.9989312029264594e-05, "loss": 4.3587, "step": 12000 }, { "epoch": 0.01, "learning_rate": 4.9988847334884796e-05, "loss": 4.3126, "step": 12500 }, { "epoch": 0.01, "learning_rate": 4.998838264050499e-05, "loss": 4.2658, "step": 13000 }, { "epoch": 0.01, "learning_rate": 4.998791794612519e-05, "loss": 4.2412, "step": 13500 }, { "epoch": 0.01, "learning_rate": 4.9987453251745395e-05, "loss": 4.2035, "step": 14000 }, { "epoch": 0.01, "learning_rate": 4.998698855736559e-05, "loss": 4.1695, "step": 14500 }, { "epoch": 0.01, "learning_rate": 4.998652386298579e-05, "loss": 4.1406, "step": 15000 }, { "epoch": 0.01, "learning_rate": 4.9986059168605994e-05, "loss": 4.1275, "step": 15500 }, { "epoch": 0.01, "learning_rate": 4.9985594474226196e-05, "loss": 4.1032, "step": 16000 }, { "epoch": 0.02, "learning_rate": 4.99851297798464e-05, "loss": 4.0588, "step": 16500 }, { "epoch": 0.02, "learning_rate": 4.9984665085466594e-05, "loss": 4.0536, "step": 17000 }, { "epoch": 0.02, "learning_rate": 4.9984200391086796e-05, "loss": 3.9979, "step": 17500 }, { "epoch": 0.02, "learning_rate": 4.998373569670699e-05, "loss": 3.9913, "step": 18000 }, { "epoch": 0.02, "learning_rate": 4.998327100232719e-05, "loss": 3.988, "step": 18500 }, { "epoch": 0.02, "learning_rate": 4.9982806307947395e-05, "loss": 3.9575, "step": 19000 }, { "epoch": 0.02, "learning_rate": 4.998234161356759e-05, "loss": 3.942, "step": 19500 }, { "epoch": 0.02, "learning_rate": 4.998187691918779e-05, "loss": 3.9258, "step": 20000 }, { "epoch": 0.02, "learning_rate": 4.998141222480799e-05, "loss": 3.8736, "step": 20500 }, { "epoch": 0.02, "learning_rate": 4.998094753042819e-05, "loss": 3.903, "step": 21000 }, { "epoch": 0.2, "learning_rate": 4.9804810945458605e-05, "loss": 3.7108, "step": 21500 }, { "epoch": 0.2, "learning_rate": 4.980016358701714e-05, "loss": 3.6276, "step": 22000 }, { "epoch": 0.21, "learning_rate": 4.979551622857568e-05, "loss": 3.568, "step": 22500 }, { "epoch": 0.21, "learning_rate": 4.9790868870134216e-05, "loss": 3.5307, "step": 23000 }, { "epoch": 0.22, "learning_rate": 4.978622151169275e-05, "loss": 3.4857, "step": 23500 }, { "epoch": 0.22, "learning_rate": 4.978157415325129e-05, "loss": 3.4463, "step": 24000 }, { "epoch": 0.23, "learning_rate": 4.977692679480983e-05, "loss": 3.4396, "step": 24500 }, { "epoch": 0.23, "learning_rate": 4.977227943636837e-05, "loss": 3.402, "step": 25000 }, { "epoch": 0.24, "learning_rate": 4.9767632077926904e-05, "loss": 3.3716, "step": 25500 }, { "epoch": 0.24, "learning_rate": 4.9762984719485446e-05, "loss": 3.3533, "step": 26000 }, { "epoch": 0.25, "learning_rate": 4.975833736104398e-05, "loss": 3.3084, "step": 26500 }, { "epoch": 0.25, "learning_rate": 4.975369000260252e-05, "loss": 3.293, "step": 27000 }, { "epoch": 0.26, "learning_rate": 4.974904264416106e-05, "loss": 3.2707, "step": 27500 }, { "epoch": 0.26, "learning_rate": 4.97443952857196e-05, "loss": 3.2561, "step": 28000 }, { "epoch": 0.26, "learning_rate": 4.9739747927278134e-05, "loss": 3.2219, "step": 28500 }, { "epoch": 0.27, "learning_rate": 4.9735100568836676e-05, "loss": 3.2014, "step": 29000 }, { "epoch": 0.27, "learning_rate": 4.973045321039521e-05, "loss": 3.1892, "step": 29500 }, { "epoch": 0.28, "learning_rate": 4.972580585195375e-05, "loss": 3.1734, "step": 30000 }, { "epoch": 0.28, "learning_rate": 4.972115849351229e-05, "loss": 3.1624, "step": 30500 }, { "epoch": 0.29, "learning_rate": 4.971651113507082e-05, "loss": 3.1416, "step": 31000 }, { "epoch": 0.29, "learning_rate": 4.9711863776629364e-05, "loss": 3.1255, "step": 31500 }, { "epoch": 0.3, "learning_rate": 4.97072164181879e-05, "loss": 3.1074, "step": 32000 }, { "epoch": 0.3, "learning_rate": 4.970256905974644e-05, "loss": 3.0992, "step": 32500 }, { "epoch": 0.31, "learning_rate": 4.9697921701304976e-05, "loss": 3.0867, "step": 33000 }, { "epoch": 0.31, "learning_rate": 4.969327434286352e-05, "loss": 3.0711, "step": 33500 }, { "epoch": 0.32, "learning_rate": 4.968862698442205e-05, "loss": 3.0621, "step": 34000 }, { "epoch": 0.32, "learning_rate": 4.9683979625980594e-05, "loss": 3.0543, "step": 34500 }, { "epoch": 0.33, "learning_rate": 4.967933226753913e-05, "loss": 3.0407, "step": 35000 }, { "epoch": 0.33, "learning_rate": 4.967468490909768e-05, "loss": 3.0363, "step": 35500 }, { "epoch": 0.33, "learning_rate": 4.967003755065621e-05, "loss": 3.0204, "step": 36000 }, { "epoch": 0.34, "learning_rate": 4.966539019221475e-05, "loss": 3.0192, "step": 36500 }, { "epoch": 0.34, "learning_rate": 4.966074283377329e-05, "loss": 3.0018, "step": 37000 }, { "epoch": 0.35, "learning_rate": 4.9656095475331824e-05, "loss": 2.992, "step": 37500 }, { "epoch": 0.35, "learning_rate": 4.9651448116890365e-05, "loss": 2.9887, "step": 38000 }, { "epoch": 0.36, "learning_rate": 4.96468007584489e-05, "loss": 2.9729, "step": 38500 }, { "epoch": 0.36, "learning_rate": 4.964215340000744e-05, "loss": 2.9743, "step": 39000 }, { "epoch": 0.37, "learning_rate": 4.963750604156598e-05, "loss": 2.9579, "step": 39500 }, { "epoch": 0.37, "learning_rate": 4.963285868312452e-05, "loss": 2.9507, "step": 40000 }, { "epoch": 0.38, "learning_rate": 4.9628211324683054e-05, "loss": 2.9553, "step": 40500 }, { "epoch": 0.38, "learning_rate": 4.9623563966241595e-05, "loss": 2.9419, "step": 41000 }, { "epoch": 0.39, "learning_rate": 4.961891660780013e-05, "loss": 2.9302, "step": 41500 }, { "epoch": 0.39, "learning_rate": 4.961426924935867e-05, "loss": 2.934, "step": 42000 }, { "epoch": 0.39, "learning_rate": 4.960962189091721e-05, "loss": 2.9192, "step": 42500 }, { "epoch": 0.4, "learning_rate": 4.960497453247575e-05, "loss": 2.9089, "step": 43000 }, { "epoch": 0.4, "learning_rate": 4.9600327174034283e-05, "loss": 2.9131, "step": 43500 }, { "epoch": 0.41, "learning_rate": 4.959567981559282e-05, "loss": 2.8963, "step": 44000 }, { "epoch": 0.41, "learning_rate": 4.959103245715136e-05, "loss": 2.8968, "step": 44500 }, { "epoch": 0.42, "learning_rate": 4.9586385098709895e-05, "loss": 2.8909, "step": 45000 }, { "epoch": 0.42, "learning_rate": 4.958173774026844e-05, "loss": 2.8808, "step": 45500 }, { "epoch": 0.43, "learning_rate": 4.957709038182697e-05, "loss": 2.879, "step": 46000 }, { "epoch": 0.43, "learning_rate": 4.957244302338551e-05, "loss": 2.8674, "step": 46500 }, { "epoch": 0.44, "learning_rate": 4.956779566494405e-05, "loss": 2.8615, "step": 47000 }, { "epoch": 0.44, "learning_rate": 4.956314830650259e-05, "loss": 2.8619, "step": 47500 }, { "epoch": 0.45, "learning_rate": 4.9558500948061125e-05, "loss": 2.8642, "step": 48000 }, { "epoch": 0.45, "learning_rate": 4.9553853589619667e-05, "loss": 2.8486, "step": 48500 }, { "epoch": 0.46, "learning_rate": 4.95492062311782e-05, "loss": 2.8469, "step": 49000 }, { "epoch": 0.46, "learning_rate": 4.954455887273674e-05, "loss": 2.8438, "step": 49500 }, { "epoch": 0.46, "learning_rate": 4.953991151429528e-05, "loss": 2.832, "step": 50000 }, { "epoch": 0.47, "learning_rate": 4.953526415585382e-05, "loss": 2.8319, "step": 50500 }, { "epoch": 0.47, "learning_rate": 4.9530616797412355e-05, "loss": 2.8332, "step": 51000 }, { "epoch": 0.48, "learning_rate": 4.952596943897089e-05, "loss": 2.8333, "step": 51500 }, { "epoch": 0.48, "learning_rate": 4.952132208052943e-05, "loss": 2.8224, "step": 52000 }, { "epoch": 0.49, "learning_rate": 4.9516674722087966e-05, "loss": 2.8138, "step": 52500 }, { "epoch": 0.49, "learning_rate": 4.951202736364651e-05, "loss": 2.8129, "step": 53000 }, { "epoch": 0.5, "learning_rate": 4.950738000520504e-05, "loss": 2.8076, "step": 53500 }, { "epoch": 0.5, "learning_rate": 4.9502732646763585e-05, "loss": 2.7981, "step": 54000 }, { "epoch": 0.51, "learning_rate": 4.949808528832212e-05, "loss": 2.8029, "step": 54500 }, { "epoch": 0.51, "learning_rate": 4.949343792988066e-05, "loss": 2.7994, "step": 55000 }, { "epoch": 0.52, "learning_rate": 4.9488790571439196e-05, "loss": 2.793, "step": 55500 }, { "epoch": 0.52, "learning_rate": 4.948414321299774e-05, "loss": 2.7945, "step": 56000 }, { "epoch": 0.53, "learning_rate": 4.947949585455627e-05, "loss": 2.7819, "step": 56500 }, { "epoch": 0.53, "learning_rate": 4.9474848496114814e-05, "loss": 2.7854, "step": 57000 }, { "epoch": 0.53, "learning_rate": 4.947020113767335e-05, "loss": 2.7698, "step": 57500 }, { "epoch": 0.54, "learning_rate": 4.946555377923189e-05, "loss": 2.7791, "step": 58000 }, { "epoch": 0.54, "learning_rate": 4.9460906420790426e-05, "loss": 2.7766, "step": 58500 }, { "epoch": 0.55, "learning_rate": 4.945625906234896e-05, "loss": 2.7627, "step": 59000 }, { "epoch": 0.55, "learning_rate": 4.94516117039075e-05, "loss": 2.7705, "step": 59500 }, { "epoch": 0.56, "learning_rate": 4.944696434546604e-05, "loss": 2.7582, "step": 60000 }, { "epoch": 0.56, "learning_rate": 4.944231698702458e-05, "loss": 2.7632, "step": 60500 }, { "epoch": 0.57, "learning_rate": 4.9437669628583114e-05, "loss": 2.764, "step": 61000 }, { "epoch": 0.57, "learning_rate": 4.9433022270141656e-05, "loss": 2.753, "step": 61500 }, { "epoch": 0.58, "learning_rate": 4.942837491170019e-05, "loss": 2.7469, "step": 62000 }, { "epoch": 0.58, "learning_rate": 4.942372755325873e-05, "loss": 2.7506, "step": 62500 }, { "epoch": 0.59, "learning_rate": 4.941908019481727e-05, "loss": 2.746, "step": 63000 }, { "epoch": 0.59, "learning_rate": 4.941443283637581e-05, "loss": 2.7373, "step": 63500 }, { "epoch": 0.59, "learning_rate": 4.9409785477934344e-05, "loss": 2.7325, "step": 64000 }, { "epoch": 0.6, "learning_rate": 4.9405138119492886e-05, "loss": 2.7329, "step": 64500 }, { "epoch": 0.6, "learning_rate": 4.940049076105142e-05, "loss": 2.7313, "step": 65000 }, { "epoch": 0.61, "learning_rate": 4.939584340260996e-05, "loss": 2.728, "step": 65500 }, { "epoch": 0.61, "learning_rate": 4.93911960441685e-05, "loss": 2.7244, "step": 66000 }, { "epoch": 0.62, "learning_rate": 4.938654868572703e-05, "loss": 2.722, "step": 66500 }, { "epoch": 0.62, "learning_rate": 4.9381901327285574e-05, "loss": 2.7278, "step": 67000 }, { "epoch": 0.63, "learning_rate": 4.937725396884411e-05, "loss": 2.7148, "step": 67500 }, { "epoch": 0.63, "learning_rate": 4.937260661040265e-05, "loss": 2.7251, "step": 68000 }, { "epoch": 0.64, "learning_rate": 4.9367959251961185e-05, "loss": 2.7076, "step": 68500 }, { "epoch": 0.64, "learning_rate": 4.936331189351973e-05, "loss": 2.717, "step": 69000 }, { "epoch": 0.65, "learning_rate": 4.935866453507826e-05, "loss": 2.7075, "step": 69500 }, { "epoch": 0.65, "learning_rate": 4.9354017176636804e-05, "loss": 2.7132, "step": 70000 }, { "epoch": 0.66, "learning_rate": 4.934936981819534e-05, "loss": 2.7055, "step": 70500 }, { "epoch": 0.66, "learning_rate": 4.934472245975388e-05, "loss": 2.7106, "step": 71000 }, { "epoch": 0.66, "learning_rate": 4.9340075101312415e-05, "loss": 2.71, "step": 71500 }, { "epoch": 0.67, "learning_rate": 4.933542774287096e-05, "loss": 2.6932, "step": 72000 }, { "epoch": 0.67, "learning_rate": 4.933078038442949e-05, "loss": 2.7055, "step": 72500 }, { "epoch": 0.68, "learning_rate": 4.9326133025988034e-05, "loss": 2.7001, "step": 73000 }, { "epoch": 0.68, "learning_rate": 4.932148566754657e-05, "loss": 2.6966, "step": 73500 }, { "epoch": 0.69, "learning_rate": 4.93168383091051e-05, "loss": 2.6817, "step": 74000 }, { "epoch": 0.69, "learning_rate": 4.9312190950663645e-05, "loss": 2.6838, "step": 74500 }, { "epoch": 0.7, "learning_rate": 4.930754359222218e-05, "loss": 2.6867, "step": 75000 }, { "epoch": 0.7, "learning_rate": 4.930289623378072e-05, "loss": 2.6786, "step": 75500 }, { "epoch": 0.71, "learning_rate": 4.929824887533926e-05, "loss": 2.6769, "step": 76000 }, { "epoch": 0.71, "learning_rate": 4.92936015168978e-05, "loss": 2.668, "step": 76500 }, { "epoch": 0.72, "learning_rate": 4.928895415845633e-05, "loss": 2.6775, "step": 77000 }, { "epoch": 0.72, "learning_rate": 4.9284306800014875e-05, "loss": 2.6715, "step": 77500 }, { "epoch": 0.72, "learning_rate": 4.927965944157341e-05, "loss": 2.67, "step": 78000 }, { "epoch": 0.73, "learning_rate": 4.927501208313195e-05, "loss": 2.6713, "step": 78500 }, { "epoch": 0.73, "learning_rate": 4.9270364724690486e-05, "loss": 2.6632, "step": 79000 }, { "epoch": 0.74, "learning_rate": 4.926571736624903e-05, "loss": 2.6631, "step": 79500 }, { "epoch": 0.74, "learning_rate": 4.926107000780756e-05, "loss": 2.6666, "step": 80000 }, { "epoch": 0.75, "learning_rate": 4.9256422649366105e-05, "loss": 2.6637, "step": 80500 }, { "epoch": 0.75, "learning_rate": 4.925177529092464e-05, "loss": 2.6602, "step": 81000 }, { "epoch": 0.76, "learning_rate": 4.924712793248318e-05, "loss": 2.657, "step": 81500 }, { "epoch": 0.76, "learning_rate": 4.9242480574041716e-05, "loss": 2.6548, "step": 82000 }, { "epoch": 0.77, "learning_rate": 4.923783321560025e-05, "loss": 2.6514, "step": 82500 }, { "epoch": 0.77, "learning_rate": 4.923318585715879e-05, "loss": 2.6485, "step": 83000 }, { "epoch": 0.78, "learning_rate": 4.922853849871733e-05, "loss": 2.6519, "step": 83500 }, { "epoch": 0.78, "learning_rate": 4.922389114027587e-05, "loss": 2.6462, "step": 84000 }, { "epoch": 0.79, "learning_rate": 4.9219243781834404e-05, "loss": 2.6415, "step": 84500 }, { "epoch": 0.79, "learning_rate": 4.9214596423392946e-05, "loss": 2.6441, "step": 85000 }, { "epoch": 0.79, "learning_rate": 4.920994906495148e-05, "loss": 2.6473, "step": 85500 }, { "epoch": 0.8, "learning_rate": 4.920530170651002e-05, "loss": 2.6343, "step": 86000 }, { "epoch": 0.8, "learning_rate": 4.920065434806856e-05, "loss": 2.634, "step": 86500 }, { "epoch": 0.81, "learning_rate": 4.91960069896271e-05, "loss": 2.6393, "step": 87000 }, { "epoch": 0.81, "learning_rate": 4.9191359631185634e-05, "loss": 2.6328, "step": 87500 }, { "epoch": 0.82, "learning_rate": 4.9186712272744176e-05, "loss": 2.6286, "step": 88000 }, { "epoch": 0.82, "learning_rate": 4.918206491430271e-05, "loss": 2.6346, "step": 88500 }, { "epoch": 0.83, "learning_rate": 4.917741755586125e-05, "loss": 2.6302, "step": 89000 }, { "epoch": 0.83, "learning_rate": 4.917277019741979e-05, "loss": 2.6394, "step": 89500 }, { "epoch": 0.84, "learning_rate": 4.916812283897832e-05, "loss": 2.6319, "step": 90000 }, { "epoch": 0.84, "learning_rate": 4.9163475480536864e-05, "loss": 2.6225, "step": 90500 }, { "epoch": 0.85, "learning_rate": 4.91588281220954e-05, "loss": 2.6239, "step": 91000 } ], "logging_steps": 500, "max_steps": 5379900, "num_input_tokens_seen": 0, "num_train_epochs": 50, "save_steps": 1000, "total_flos": 1.09831567835136e+19, "train_batch_size": 7, "trial_name": null, "trial_params": null }