{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 8000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00125, "grad_norm": 1.0625, "learning_rate": 4.05e-06, "loss": 2.85, "step": 10 }, { "epoch": 0.0025, "grad_norm": 1.078125, "learning_rate": 8.550000000000001e-06, "loss": 2.9, "step": 20 }, { "epoch": 0.00375, "grad_norm": 1.1171875, "learning_rate": 1.305e-05, "loss": 2.825, "step": 30 }, { "epoch": 0.005, "grad_norm": 0.94140625, "learning_rate": 1.7550000000000003e-05, "loss": 2.8625, "step": 40 }, { "epoch": 0.00625, "grad_norm": 0.8828125, "learning_rate": 2.205e-05, "loss": 2.825, "step": 50 }, { "epoch": 0.0075, "grad_norm": 0.796875, "learning_rate": 2.655e-05, "loss": 2.7875, "step": 60 }, { "epoch": 0.00875, "grad_norm": 0.8046875, "learning_rate": 3.105e-05, "loss": 2.8125, "step": 70 }, { "epoch": 0.01, "grad_norm": 0.828125, "learning_rate": 3.5550000000000004e-05, "loss": 2.75, "step": 80 }, { "epoch": 0.01125, "grad_norm": 0.72265625, "learning_rate": 4.0050000000000004e-05, "loss": 2.7375, "step": 90 }, { "epoch": 0.0125, "grad_norm": 0.7578125, "learning_rate": 4.4550000000000005e-05, "loss": 2.775, "step": 100 }, { "epoch": 0.01375, "grad_norm": 0.79296875, "learning_rate": 4.9050000000000006e-05, "loss": 2.725, "step": 110 }, { "epoch": 0.015, "grad_norm": 0.75, "learning_rate": 5.355e-05, "loss": 2.7125, "step": 120 }, { "epoch": 0.01625, "grad_norm": 0.71484375, "learning_rate": 5.805000000000001e-05, "loss": 2.8, "step": 130 }, { "epoch": 0.0175, "grad_norm": 0.734375, "learning_rate": 6.255e-05, "loss": 2.7, "step": 140 }, { "epoch": 0.01875, "grad_norm": 0.7109375, "learning_rate": 6.705e-05, "loss": 2.725, "step": 150 }, { "epoch": 0.02, "grad_norm": 0.75390625, "learning_rate": 7.155e-05, "loss": 2.675, "step": 160 }, { "epoch": 0.02125, "grad_norm": 0.7421875, "learning_rate": 7.605e-05, "loss": 2.6375, "step": 170 }, { "epoch": 0.0225, "grad_norm": 0.71875, "learning_rate": 8.055e-05, "loss": 2.7, "step": 180 }, { "epoch": 0.02375, "grad_norm": 0.7265625, "learning_rate": 8.505000000000001e-05, "loss": 2.6875, "step": 190 }, { "epoch": 0.025, "grad_norm": 0.71484375, "learning_rate": 8.955000000000001e-05, "loss": 2.725, "step": 200 }, { "epoch": 0.02625, "grad_norm": 0.72265625, "learning_rate": 8.999970435019192e-05, "loss": 2.675, "step": 210 }, { "epoch": 0.0275, "grad_norm": 0.7890625, "learning_rate": 8.99986823558428e-05, "loss": 2.7, "step": 220 }, { "epoch": 0.02875, "grad_norm": 0.73046875, "learning_rate": 8.999693038353013e-05, "loss": 2.7375, "step": 230 }, { "epoch": 0.03, "grad_norm": 0.7421875, "learning_rate": 8.999444846167473e-05, "loss": 2.6875, "step": 240 }, { "epoch": 0.03125, "grad_norm": 0.75390625, "learning_rate": 8.999123663053887e-05, "loss": 2.675, "step": 250 }, { "epoch": 0.0325, "grad_norm": 0.66796875, "learning_rate": 8.998729494222554e-05, "loss": 2.7375, "step": 260 }, { "epoch": 0.03375, "grad_norm": 0.74609375, "learning_rate": 8.998262346067761e-05, "loss": 2.675, "step": 270 }, { "epoch": 0.035, "grad_norm": 0.66796875, "learning_rate": 8.997722226167686e-05, "loss": 2.725, "step": 280 }, { "epoch": 0.03625, "grad_norm": 0.71484375, "learning_rate": 8.997109143284262e-05, "loss": 2.65, "step": 290 }, { "epoch": 0.0375, "grad_norm": 0.7109375, "learning_rate": 8.996423107363048e-05, "loss": 2.7125, "step": 300 }, { "epoch": 0.03875, "grad_norm": 0.6875, "learning_rate": 8.995664129533062e-05, "loss": 2.675, "step": 310 }, { "epoch": 0.04, "grad_norm": 0.6953125, "learning_rate": 8.994832222106598e-05, "loss": 2.65, "step": 320 }, { "epoch": 0.04125, "grad_norm": 0.69921875, "learning_rate": 8.993927398579033e-05, "loss": 2.6375, "step": 330 }, { "epoch": 0.0425, "grad_norm": 0.6796875, "learning_rate": 8.992949673628603e-05, "loss": 2.625, "step": 340 }, { "epoch": 0.04375, "grad_norm": 0.6640625, "learning_rate": 8.99189906311616e-05, "loss": 2.7375, "step": 350 }, { "epoch": 0.045, "grad_norm": 0.69921875, "learning_rate": 8.990775584084932e-05, "loss": 2.675, "step": 360 }, { "epoch": 0.04625, "grad_norm": 0.70703125, "learning_rate": 8.989579254760225e-05, "loss": 2.7, "step": 370 }, { "epoch": 0.0475, "grad_norm": 0.67578125, "learning_rate": 8.988310094549142e-05, "loss": 2.675, "step": 380 }, { "epoch": 0.04875, "grad_norm": 0.66796875, "learning_rate": 8.986968124040265e-05, "loss": 2.7, "step": 390 }, { "epoch": 0.05, "grad_norm": 0.65234375, "learning_rate": 8.985553365003315e-05, "loss": 2.675, "step": 400 }, { "epoch": 0.05125, "grad_norm": 0.69140625, "learning_rate": 8.984065840388811e-05, "loss": 2.725, "step": 410 }, { "epoch": 0.0525, "grad_norm": 0.703125, "learning_rate": 8.982505574327683e-05, "loss": 2.65, "step": 420 }, { "epoch": 0.05375, "grad_norm": 0.68359375, "learning_rate": 8.980872592130892e-05, "loss": 2.725, "step": 430 }, { "epoch": 0.055, "grad_norm": 0.66015625, "learning_rate": 8.979166920289013e-05, "loss": 2.7375, "step": 440 }, { "epoch": 0.05625, "grad_norm": 0.70703125, "learning_rate": 8.97738858647181e-05, "loss": 2.7625, "step": 450 }, { "epoch": 0.0575, "grad_norm": 0.734375, "learning_rate": 8.975537619527783e-05, "loss": 2.7, "step": 460 }, { "epoch": 0.05875, "grad_norm": 0.64453125, "learning_rate": 8.973614049483699e-05, "loss": 2.6625, "step": 470 }, { "epoch": 0.06, "grad_norm": 0.69921875, "learning_rate": 8.971617907544115e-05, "loss": 2.6875, "step": 480 }, { "epoch": 0.06125, "grad_norm": 0.73046875, "learning_rate": 8.969549226090858e-05, "loss": 2.6375, "step": 490 }, { "epoch": 0.0625, "grad_norm": 0.6953125, "learning_rate": 8.967408038682505e-05, "loss": 2.6625, "step": 500 }, { "epoch": 0.06375, "grad_norm": 0.65625, "learning_rate": 8.965194380053847e-05, "loss": 2.675, "step": 510 }, { "epoch": 0.065, "grad_norm": 0.7265625, "learning_rate": 8.962908286115311e-05, "loss": 2.6875, "step": 520 }, { "epoch": 0.06625, "grad_norm": 0.7109375, "learning_rate": 8.960549793952388e-05, "loss": 2.6625, "step": 530 }, { "epoch": 0.0675, "grad_norm": 0.7109375, "learning_rate": 8.958118941825026e-05, "loss": 2.7, "step": 540 }, { "epoch": 0.06875, "grad_norm": 0.7734375, "learning_rate": 8.955615769167015e-05, "loss": 2.7125, "step": 550 }, { "epoch": 0.07, "grad_norm": 0.6875, "learning_rate": 8.953040316585341e-05, "loss": 2.6375, "step": 560 }, { "epoch": 0.07125, "grad_norm": 0.65234375, "learning_rate": 8.95039262585953e-05, "loss": 2.75, "step": 570 }, { "epoch": 0.0725, "grad_norm": 0.68359375, "learning_rate": 8.947672739940968e-05, "loss": 2.6375, "step": 580 }, { "epoch": 0.07375, "grad_norm": 0.65234375, "learning_rate": 8.944880702952212e-05, "loss": 2.675, "step": 590 }, { "epoch": 0.075, "grad_norm": 0.66015625, "learning_rate": 8.942016560186262e-05, "loss": 2.6875, "step": 600 }, { "epoch": 0.07625, "grad_norm": 0.6875, "learning_rate": 8.939080358105841e-05, "loss": 2.725, "step": 610 }, { "epoch": 0.0775, "grad_norm": 0.66015625, "learning_rate": 8.936072144342625e-05, "loss": 2.675, "step": 620 }, { "epoch": 0.07875, "grad_norm": 0.6953125, "learning_rate": 8.932991967696483e-05, "loss": 2.725, "step": 630 }, { "epoch": 0.08, "grad_norm": 0.6640625, "learning_rate": 8.929839878134681e-05, "loss": 2.725, "step": 640 }, { "epoch": 0.08125, "grad_norm": 0.6875, "learning_rate": 8.926615926791068e-05, "loss": 2.7125, "step": 650 }, { "epoch": 0.0825, "grad_norm": 0.71875, "learning_rate": 8.923320165965251e-05, "loss": 2.6625, "step": 660 }, { "epoch": 0.08375, "grad_norm": 0.703125, "learning_rate": 8.919952649121748e-05, "loss": 2.725, "step": 670 }, { "epoch": 0.085, "grad_norm": 0.671875, "learning_rate": 8.916513430889116e-05, "loss": 2.625, "step": 680 }, { "epoch": 0.08625, "grad_norm": 0.69921875, "learning_rate": 8.913002567059068e-05, "loss": 2.675, "step": 690 }, { "epoch": 0.0875, "grad_norm": 0.640625, "learning_rate": 8.909420114585561e-05, "loss": 2.675, "step": 700 }, { "epoch": 0.08875, "grad_norm": 0.66796875, "learning_rate": 8.90576613158389e-05, "loss": 2.7, "step": 710 }, { "epoch": 0.09, "grad_norm": 0.6875, "learning_rate": 8.902040677329717e-05, "loss": 2.6875, "step": 720 }, { "epoch": 0.09125, "grad_norm": 0.72265625, "learning_rate": 8.898243812258141e-05, "loss": 2.6625, "step": 730 }, { "epoch": 0.0925, "grad_norm": 0.671875, "learning_rate": 8.894375597962691e-05, "loss": 2.625, "step": 740 }, { "epoch": 0.09375, "grad_norm": 0.70703125, "learning_rate": 8.890436097194343e-05, "loss": 2.6875, "step": 750 }, { "epoch": 0.095, "grad_norm": 0.66796875, "learning_rate": 8.886425373860498e-05, "loss": 2.6875, "step": 760 }, { "epoch": 0.09625, "grad_norm": 0.6796875, "learning_rate": 8.88234349302394e-05, "loss": 2.7375, "step": 770 }, { "epoch": 0.0975, "grad_norm": 0.7109375, "learning_rate": 8.878190520901788e-05, "loss": 2.6375, "step": 780 }, { "epoch": 0.09875, "grad_norm": 0.70703125, "learning_rate": 8.873966524864417e-05, "loss": 2.6875, "step": 790 }, { "epoch": 0.1, "grad_norm": 0.69140625, "learning_rate": 8.869671573434369e-05, "loss": 2.675, "step": 800 }, { "epoch": 0.10125, "grad_norm": 0.70703125, "learning_rate": 8.865305736285234e-05, "loss": 2.6875, "step": 810 }, { "epoch": 0.1025, "grad_norm": 0.67578125, "learning_rate": 8.860869084240533e-05, "loss": 2.625, "step": 820 }, { "epoch": 0.10375, "grad_norm": 0.6640625, "learning_rate": 8.85636168927255e-05, "loss": 2.675, "step": 830 }, { "epoch": 0.105, "grad_norm": 0.7109375, "learning_rate": 8.851783624501188e-05, "loss": 2.6625, "step": 840 }, { "epoch": 0.10625, "grad_norm": 0.6484375, "learning_rate": 8.84713496419276e-05, "loss": 2.6875, "step": 850 }, { "epoch": 0.1075, "grad_norm": 0.625, "learning_rate": 8.842415783758798e-05, "loss": 2.65, "step": 860 }, { "epoch": 0.10875, "grad_norm": 0.65625, "learning_rate": 8.837626159754828e-05, "loss": 2.6875, "step": 870 }, { "epoch": 0.11, "grad_norm": 0.73828125, "learning_rate": 8.832766169879124e-05, "loss": 2.6875, "step": 880 }, { "epoch": 0.11125, "grad_norm": 0.6640625, "learning_rate": 8.827835892971452e-05, "loss": 2.6875, "step": 890 }, { "epoch": 0.1125, "grad_norm": 0.72265625, "learning_rate": 8.822835409011783e-05, "loss": 2.7, "step": 900 }, { "epoch": 0.11375, "grad_norm": 0.6484375, "learning_rate": 8.81776479911901e-05, "loss": 2.7125, "step": 910 }, { "epoch": 0.115, "grad_norm": 0.671875, "learning_rate": 8.812624145549617e-05, "loss": 2.675, "step": 920 }, { "epoch": 0.11625, "grad_norm": 0.734375, "learning_rate": 8.807413531696352e-05, "loss": 2.625, "step": 930 }, { "epoch": 0.1175, "grad_norm": 0.78125, "learning_rate": 8.802133042086876e-05, "loss": 2.7125, "step": 940 }, { "epoch": 0.11875, "grad_norm": 0.6484375, "learning_rate": 8.796782762382386e-05, "loss": 2.7, "step": 950 }, { "epoch": 0.12, "grad_norm": 0.6875, "learning_rate": 8.791362779376231e-05, "loss": 2.725, "step": 960 }, { "epoch": 0.12125, "grad_norm": 0.6875, "learning_rate": 8.7858731809925e-05, "loss": 2.7, "step": 970 }, { "epoch": 0.1225, "grad_norm": 0.67578125, "learning_rate": 8.780314056284595e-05, "loss": 2.65, "step": 980 }, { "epoch": 0.12375, "grad_norm": 0.66796875, "learning_rate": 8.774685495433793e-05, "loss": 2.6375, "step": 990 }, { "epoch": 0.125, "grad_norm": 0.69921875, "learning_rate": 8.768987589747775e-05, "loss": 2.725, "step": 1000 }, { "epoch": 0.12625, "grad_norm": 0.66796875, "learning_rate": 8.763220431659151e-05, "loss": 2.6625, "step": 1010 }, { "epoch": 0.1275, "grad_norm": 0.62890625, "learning_rate": 8.757384114723954e-05, "loss": 2.65, "step": 1020 }, { "epoch": 0.12875, "grad_norm": 0.69921875, "learning_rate": 8.751478733620131e-05, "loss": 2.7, "step": 1030 }, { "epoch": 0.13, "grad_norm": 0.66015625, "learning_rate": 8.745504384145999e-05, "loss": 2.675, "step": 1040 }, { "epoch": 0.13125, "grad_norm": 0.66796875, "learning_rate": 8.739461163218694e-05, "loss": 2.6625, "step": 1050 }, { "epoch": 0.1325, "grad_norm": 0.65625, "learning_rate": 8.733349168872599e-05, "loss": 2.65, "step": 1060 }, { "epoch": 0.13375, "grad_norm": 0.69921875, "learning_rate": 8.72716850025776e-05, "loss": 2.7125, "step": 1070 }, { "epoch": 0.135, "grad_norm": 0.625, "learning_rate": 8.720919257638259e-05, "loss": 2.6875, "step": 1080 }, { "epoch": 0.13625, "grad_norm": 0.67578125, "learning_rate": 8.714601542390611e-05, "loss": 2.65, "step": 1090 }, { "epoch": 0.1375, "grad_norm": 0.60546875, "learning_rate": 8.708215457002103e-05, "loss": 2.7125, "step": 1100 }, { "epoch": 0.13875, "grad_norm": 0.65234375, "learning_rate": 8.701761105069137e-05, "loss": 2.725, "step": 1110 }, { "epoch": 0.14, "grad_norm": 0.61328125, "learning_rate": 8.69523859129555e-05, "loss": 2.675, "step": 1120 }, { "epoch": 0.14125, "grad_norm": 0.7109375, "learning_rate": 8.688648021490913e-05, "loss": 2.675, "step": 1130 }, { "epoch": 0.1425, "grad_norm": 0.69921875, "learning_rate": 8.681989502568818e-05, "loss": 2.6625, "step": 1140 }, { "epoch": 0.14375, "grad_norm": 0.65234375, "learning_rate": 8.675263142545141e-05, "loss": 2.65, "step": 1150 }, { "epoch": 0.145, "grad_norm": 0.65234375, "learning_rate": 8.668469050536289e-05, "loss": 2.675, "step": 1160 }, { "epoch": 0.14625, "grad_norm": 0.7421875, "learning_rate": 8.661607336757433e-05, "loss": 2.75, "step": 1170 }, { "epoch": 0.1475, "grad_norm": 0.6328125, "learning_rate": 8.65467811252072e-05, "loss": 2.6125, "step": 1180 }, { "epoch": 0.14875, "grad_norm": 0.6484375, "learning_rate": 8.647681490233461e-05, "loss": 2.6375, "step": 1190 }, { "epoch": 0.15, "grad_norm": 0.66015625, "learning_rate": 8.640617583396315e-05, "loss": 2.65, "step": 1200 }, { "epoch": 0.15125, "grad_norm": 0.66015625, "learning_rate": 8.633486506601444e-05, "loss": 2.7125, "step": 1210 }, { "epoch": 0.1525, "grad_norm": 0.62890625, "learning_rate": 8.626288375530659e-05, "loss": 2.6625, "step": 1220 }, { "epoch": 0.15375, "grad_norm": 0.71875, "learning_rate": 8.619023306953534e-05, "loss": 2.625, "step": 1230 }, { "epoch": 0.155, "grad_norm": 0.71484375, "learning_rate": 8.611691418725518e-05, "loss": 2.6625, "step": 1240 }, { "epoch": 0.15625, "grad_norm": 0.609375, "learning_rate": 8.604292829786024e-05, "loss": 2.6125, "step": 1250 }, { "epoch": 0.1575, "grad_norm": 0.67578125, "learning_rate": 8.596827660156498e-05, "loss": 2.675, "step": 1260 }, { "epoch": 0.15875, "grad_norm": 0.65234375, "learning_rate": 8.58929603093847e-05, "loss": 2.6625, "step": 1270 }, { "epoch": 0.16, "grad_norm": 0.6640625, "learning_rate": 8.581698064311593e-05, "loss": 2.65, "step": 1280 }, { "epoch": 0.16125, "grad_norm": 0.6328125, "learning_rate": 8.574033883531658e-05, "loss": 2.6625, "step": 1290 }, { "epoch": 0.1625, "grad_norm": 0.65625, "learning_rate": 8.566303612928596e-05, "loss": 2.65, "step": 1300 }, { "epoch": 0.16375, "grad_norm": 0.65625, "learning_rate": 8.558507377904462e-05, "loss": 2.65, "step": 1310 }, { "epoch": 0.165, "grad_norm": 0.66015625, "learning_rate": 8.550645304931397e-05, "loss": 2.6625, "step": 1320 }, { "epoch": 0.16625, "grad_norm": 0.6328125, "learning_rate": 8.542717521549582e-05, "loss": 2.6375, "step": 1330 }, { "epoch": 0.1675, "grad_norm": 0.671875, "learning_rate": 8.53472415636516e-05, "loss": 2.675, "step": 1340 }, { "epoch": 0.16875, "grad_norm": 0.625, "learning_rate": 8.526665339048167e-05, "loss": 2.6875, "step": 1350 }, { "epoch": 0.17, "grad_norm": 0.6875, "learning_rate": 8.518541200330404e-05, "loss": 2.725, "step": 1360 }, { "epoch": 0.17125, "grad_norm": 0.62109375, "learning_rate": 8.510351872003337e-05, "loss": 2.675, "step": 1370 }, { "epoch": 0.1725, "grad_norm": 0.63671875, "learning_rate": 8.502097486915951e-05, "loss": 2.7, "step": 1380 }, { "epoch": 0.17375, "grad_norm": 0.66796875, "learning_rate": 8.493778178972592e-05, "loss": 2.6625, "step": 1390 }, { "epoch": 0.175, "grad_norm": 0.6015625, "learning_rate": 8.485394083130802e-05, "loss": 2.65, "step": 1400 }, { "epoch": 0.17625, "grad_norm": 0.66015625, "learning_rate": 8.476945335399122e-05, "loss": 2.6875, "step": 1410 }, { "epoch": 0.1775, "grad_norm": 0.65234375, "learning_rate": 8.46843207283489e-05, "loss": 2.7125, "step": 1420 }, { "epoch": 0.17875, "grad_norm": 0.65234375, "learning_rate": 8.459854433542021e-05, "loss": 2.6375, "step": 1430 }, { "epoch": 0.18, "grad_norm": 0.61328125, "learning_rate": 8.451212556668756e-05, "loss": 2.65, "step": 1440 }, { "epoch": 0.18125, "grad_norm": 0.671875, "learning_rate": 8.44250658240542e-05, "loss": 2.6, "step": 1450 }, { "epoch": 0.1825, "grad_norm": 0.6484375, "learning_rate": 8.433736651982129e-05, "loss": 2.6625, "step": 1460 }, { "epoch": 0.18375, "grad_norm": 0.640625, "learning_rate": 8.424902907666517e-05, "loss": 2.6625, "step": 1470 }, { "epoch": 0.185, "grad_norm": 0.640625, "learning_rate": 8.41600549276142e-05, "loss": 2.7125, "step": 1480 }, { "epoch": 0.18625, "grad_norm": 0.703125, "learning_rate": 8.407044551602544e-05, "loss": 2.6375, "step": 1490 }, { "epoch": 0.1875, "grad_norm": 0.64453125, "learning_rate": 8.398020229556138e-05, "loss": 2.7125, "step": 1500 }, { "epoch": 0.18875, "grad_norm": 0.67578125, "learning_rate": 8.388932673016629e-05, "loss": 2.625, "step": 1510 }, { "epoch": 0.19, "grad_norm": 0.68359375, "learning_rate": 8.37978202940424e-05, "loss": 2.7, "step": 1520 }, { "epoch": 0.19125, "grad_norm": 0.62890625, "learning_rate": 8.370568447162619e-05, "loss": 2.7, "step": 1530 }, { "epoch": 0.1925, "grad_norm": 0.703125, "learning_rate": 8.361292075756401e-05, "loss": 2.6125, "step": 1540 }, { "epoch": 0.19375, "grad_norm": 0.72265625, "learning_rate": 8.351953065668812e-05, "loss": 2.675, "step": 1550 }, { "epoch": 0.195, "grad_norm": 0.6953125, "learning_rate": 8.342551568399213e-05, "loss": 2.6625, "step": 1560 }, { "epoch": 0.19625, "grad_norm": 0.65234375, "learning_rate": 8.33308773646064e-05, "loss": 2.6875, "step": 1570 }, { "epoch": 0.1975, "grad_norm": 0.69140625, "learning_rate": 8.323561723377341e-05, "loss": 2.6875, "step": 1580 }, { "epoch": 0.19875, "grad_norm": 0.62109375, "learning_rate": 8.313973683682276e-05, "loss": 2.6375, "step": 1590 }, { "epoch": 0.2, "grad_norm": 0.6953125, "learning_rate": 8.304323772914617e-05, "loss": 2.6875, "step": 1600 }, { "epoch": 0.20125, "grad_norm": 0.640625, "learning_rate": 8.294612147617215e-05, "loss": 2.65, "step": 1610 }, { "epoch": 0.2025, "grad_norm": 0.7265625, "learning_rate": 8.284838965334077e-05, "loss": 2.6375, "step": 1620 }, { "epoch": 0.20375, "grad_norm": 0.6328125, "learning_rate": 8.275004384607788e-05, "loss": 2.6625, "step": 1630 }, { "epoch": 0.205, "grad_norm": 0.625, "learning_rate": 8.265108564976962e-05, "loss": 2.65, "step": 1640 }, { "epoch": 0.20625, "grad_norm": 0.65234375, "learning_rate": 8.255151666973635e-05, "loss": 2.675, "step": 1650 }, { "epoch": 0.2075, "grad_norm": 0.64453125, "learning_rate": 8.245133852120675e-05, "loss": 2.6625, "step": 1660 }, { "epoch": 0.20875, "grad_norm": 0.62109375, "learning_rate": 8.235055282929154e-05, "loss": 2.675, "step": 1670 }, { "epoch": 0.21, "grad_norm": 0.62890625, "learning_rate": 8.224916122895711e-05, "loss": 2.625, "step": 1680 }, { "epoch": 0.21125, "grad_norm": 0.62109375, "learning_rate": 8.214716536499908e-05, "loss": 2.6875, "step": 1690 }, { "epoch": 0.2125, "grad_norm": 0.65625, "learning_rate": 8.204456689201546e-05, "loss": 2.6875, "step": 1700 }, { "epoch": 0.21375, "grad_norm": 0.65234375, "learning_rate": 8.194136747438001e-05, "loss": 2.675, "step": 1710 }, { "epoch": 0.215, "grad_norm": 0.6328125, "learning_rate": 8.183756878621512e-05, "loss": 2.65, "step": 1720 }, { "epoch": 0.21625, "grad_norm": 0.66015625, "learning_rate": 8.17331725113646e-05, "loss": 2.6125, "step": 1730 }, { "epoch": 0.2175, "grad_norm": 0.69921875, "learning_rate": 8.162818034336651e-05, "loss": 2.65, "step": 1740 }, { "epoch": 0.21875, "grad_norm": 0.625, "learning_rate": 8.152259398542561e-05, "loss": 2.675, "step": 1750 }, { "epoch": 0.22, "grad_norm": 0.59375, "learning_rate": 8.141641515038566e-05, "loss": 2.675, "step": 1760 }, { "epoch": 0.22125, "grad_norm": 0.65234375, "learning_rate": 8.130964556070181e-05, "loss": 2.625, "step": 1770 }, { "epoch": 0.2225, "grad_norm": 0.6640625, "learning_rate": 8.120228694841246e-05, "loss": 2.6375, "step": 1780 }, { "epoch": 0.22375, "grad_norm": 0.671875, "learning_rate": 8.109434105511132e-05, "loss": 2.65, "step": 1790 }, { "epoch": 0.225, "grad_norm": 0.671875, "learning_rate": 8.098580963191908e-05, "loss": 2.7125, "step": 1800 }, { "epoch": 0.22625, "grad_norm": 0.6640625, "learning_rate": 8.087669443945499e-05, "loss": 2.6375, "step": 1810 }, { "epoch": 0.2275, "grad_norm": 0.65625, "learning_rate": 8.076699724780836e-05, "loss": 2.675, "step": 1820 }, { "epoch": 0.22875, "grad_norm": 0.66015625, "learning_rate": 8.065671983650978e-05, "loss": 2.625, "step": 1830 }, { "epoch": 0.23, "grad_norm": 0.640625, "learning_rate": 8.054586399450232e-05, "loss": 2.6375, "step": 1840 }, { "epoch": 0.23125, "grad_norm": 0.7109375, "learning_rate": 8.043443152011244e-05, "loss": 2.6625, "step": 1850 }, { "epoch": 0.2325, "grad_norm": 0.6328125, "learning_rate": 8.03224242210209e-05, "loss": 2.65, "step": 1860 }, { "epoch": 0.23375, "grad_norm": 0.6328125, "learning_rate": 8.02098439142333e-05, "loss": 2.6125, "step": 1870 }, { "epoch": 0.235, "grad_norm": 0.62109375, "learning_rate": 8.009669242605077e-05, "loss": 2.7, "step": 1880 }, { "epoch": 0.23625, "grad_norm": 0.6484375, "learning_rate": 7.998297159204023e-05, "loss": 2.6875, "step": 1890 }, { "epoch": 0.2375, "grad_norm": 0.625, "learning_rate": 7.986868325700463e-05, "loss": 2.6375, "step": 1900 }, { "epoch": 0.23875, "grad_norm": 0.63671875, "learning_rate": 7.975382927495309e-05, "loss": 2.6625, "step": 1910 }, { "epoch": 0.24, "grad_norm": 0.6640625, "learning_rate": 7.963841150907073e-05, "loss": 2.65, "step": 1920 }, { "epoch": 0.24125, "grad_norm": 0.6796875, "learning_rate": 7.952243183168849e-05, "loss": 2.6375, "step": 1930 }, { "epoch": 0.2425, "grad_norm": 0.6640625, "learning_rate": 7.940589212425277e-05, "loss": 2.65, "step": 1940 }, { "epoch": 0.24375, "grad_norm": 0.60546875, "learning_rate": 7.92887942772949e-05, "loss": 2.625, "step": 1950 }, { "epoch": 0.245, "grad_norm": 0.6484375, "learning_rate": 7.917114019040043e-05, "loss": 2.6375, "step": 1960 }, { "epoch": 0.24625, "grad_norm": 0.66015625, "learning_rate": 7.90529317721784e-05, "loss": 2.6375, "step": 1970 }, { "epoch": 0.2475, "grad_norm": 0.64453125, "learning_rate": 7.89341709402303e-05, "loss": 2.7, "step": 1980 }, { "epoch": 0.24875, "grad_norm": 0.6484375, "learning_rate": 7.881485962111898e-05, "loss": 2.625, "step": 1990 }, { "epoch": 0.25, "grad_norm": 0.6171875, "learning_rate": 7.869499975033741e-05, "loss": 2.75, "step": 2000 }, { "epoch": 0.25125, "grad_norm": 0.6171875, "learning_rate": 7.85745932722773e-05, "loss": 2.625, "step": 2010 }, { "epoch": 0.2525, "grad_norm": 0.640625, "learning_rate": 7.845364214019753e-05, "loss": 2.65, "step": 2020 }, { "epoch": 0.25375, "grad_norm": 0.6328125, "learning_rate": 7.833214831619247e-05, "loss": 2.6, "step": 2030 }, { "epoch": 0.255, "grad_norm": 0.6328125, "learning_rate": 7.821011377116012e-05, "loss": 2.65, "step": 2040 }, { "epoch": 0.25625, "grad_norm": 0.66015625, "learning_rate": 7.808754048477022e-05, "loss": 2.675, "step": 2050 }, { "epoch": 0.2575, "grad_norm": 0.63671875, "learning_rate": 7.796443044543205e-05, "loss": 2.6375, "step": 2060 }, { "epoch": 0.25875, "grad_norm": 0.6171875, "learning_rate": 7.78407856502622e-05, "loss": 2.5875, "step": 2070 }, { "epoch": 0.26, "grad_norm": 0.62109375, "learning_rate": 7.771660810505223e-05, "loss": 2.625, "step": 2080 }, { "epoch": 0.26125, "grad_norm": 0.6328125, "learning_rate": 7.759189982423603e-05, "loss": 2.6875, "step": 2090 }, { "epoch": 0.2625, "grad_norm": 0.65625, "learning_rate": 7.746666283085721e-05, "loss": 2.5875, "step": 2100 }, { "epoch": 0.26375, "grad_norm": 0.6015625, "learning_rate": 7.734089915653627e-05, "loss": 2.6375, "step": 2110 }, { "epoch": 0.265, "grad_norm": 0.63671875, "learning_rate": 7.721461084143765e-05, "loss": 2.65, "step": 2120 }, { "epoch": 0.26625, "grad_norm": 0.6640625, "learning_rate": 7.708779993423656e-05, "loss": 2.675, "step": 2130 }, { "epoch": 0.2675, "grad_norm": 0.6484375, "learning_rate": 7.69604684920859e-05, "loss": 2.625, "step": 2140 }, { "epoch": 0.26875, "grad_norm": 0.66796875, "learning_rate": 7.683261858058272e-05, "loss": 2.6375, "step": 2150 }, { "epoch": 0.27, "grad_norm": 0.6328125, "learning_rate": 7.670425227373484e-05, "loss": 2.6625, "step": 2160 }, { "epoch": 0.27125, "grad_norm": 0.6640625, "learning_rate": 7.657537165392712e-05, "loss": 2.7, "step": 2170 }, { "epoch": 0.2725, "grad_norm": 0.6171875, "learning_rate": 7.644597881188772e-05, "loss": 2.7125, "step": 2180 }, { "epoch": 0.27375, "grad_norm": 0.66015625, "learning_rate": 7.631607584665414e-05, "loss": 2.625, "step": 2190 }, { "epoch": 0.275, "grad_norm": 0.6640625, "learning_rate": 7.618566486553931e-05, "loss": 2.625, "step": 2200 }, { "epoch": 0.27625, "grad_norm": 0.671875, "learning_rate": 7.605474798409717e-05, "loss": 2.6625, "step": 2210 }, { "epoch": 0.2775, "grad_norm": 0.6875, "learning_rate": 7.592332732608858e-05, "loss": 2.6375, "step": 2220 }, { "epoch": 0.27875, "grad_norm": 0.6484375, "learning_rate": 7.57914050234467e-05, "loss": 2.7, "step": 2230 }, { "epoch": 0.28, "grad_norm": 0.640625, "learning_rate": 7.565898321624254e-05, "loss": 2.65, "step": 2240 }, { "epoch": 0.28125, "grad_norm": 0.65234375, "learning_rate": 7.552606405265007e-05, "loss": 2.7, "step": 2250 }, { "epoch": 0.2825, "grad_norm": 0.6328125, "learning_rate": 7.53926496889116e-05, "loss": 2.7, "step": 2260 }, { "epoch": 0.28375, "grad_norm": 0.6875, "learning_rate": 7.525874228930262e-05, "loss": 2.6875, "step": 2270 }, { "epoch": 0.285, "grad_norm": 0.62890625, "learning_rate": 7.512434402609672e-05, "loss": 2.675, "step": 2280 }, { "epoch": 0.28625, "grad_norm": 0.70703125, "learning_rate": 7.498945707953046e-05, "loss": 2.675, "step": 2290 }, { "epoch": 0.2875, "grad_norm": 0.67578125, "learning_rate": 7.485408363776787e-05, "loss": 2.6375, "step": 2300 }, { "epoch": 0.28875, "grad_norm": 0.6484375, "learning_rate": 7.471822589686499e-05, "loss": 2.6125, "step": 2310 }, { "epoch": 0.29, "grad_norm": 0.6640625, "learning_rate": 7.458188606073432e-05, "loss": 2.6875, "step": 2320 }, { "epoch": 0.29125, "grad_norm": 0.609375, "learning_rate": 7.444506634110894e-05, "loss": 2.6375, "step": 2330 }, { "epoch": 0.2925, "grad_norm": 0.64453125, "learning_rate": 7.430776895750678e-05, "loss": 2.65, "step": 2340 }, { "epoch": 0.29375, "grad_norm": 0.625, "learning_rate": 7.416999613719446e-05, "loss": 2.675, "step": 2350 }, { "epoch": 0.295, "grad_norm": 0.65234375, "learning_rate": 7.403175011515126e-05, "loss": 2.7, "step": 2360 }, { "epoch": 0.29625, "grad_norm": 0.703125, "learning_rate": 7.389303313403286e-05, "loss": 2.675, "step": 2370 }, { "epoch": 0.2975, "grad_norm": 0.62890625, "learning_rate": 7.375384744413488e-05, "loss": 2.6375, "step": 2380 }, { "epoch": 0.29875, "grad_norm": 0.671875, "learning_rate": 7.361419530335651e-05, "loss": 2.6625, "step": 2390 }, { "epoch": 0.3, "grad_norm": 0.6796875, "learning_rate": 7.347407897716373e-05, "loss": 2.6625, "step": 2400 }, { "epoch": 0.30125, "grad_norm": 0.65234375, "learning_rate": 7.333350073855267e-05, "loss": 2.6875, "step": 2410 }, { "epoch": 0.3025, "grad_norm": 0.70703125, "learning_rate": 7.319246286801269e-05, "loss": 2.625, "step": 2420 }, { "epoch": 0.30375, "grad_norm": 0.640625, "learning_rate": 7.30509676534894e-05, "loss": 2.6875, "step": 2430 }, { "epoch": 0.305, "grad_norm": 0.6640625, "learning_rate": 7.290901739034753e-05, "loss": 2.7, "step": 2440 }, { "epoch": 0.30625, "grad_norm": 0.61328125, "learning_rate": 7.27666143813337e-05, "loss": 2.6375, "step": 2450 }, { "epoch": 0.3075, "grad_norm": 0.64453125, "learning_rate": 7.262376093653907e-05, "loss": 2.625, "step": 2460 }, { "epoch": 0.30875, "grad_norm": 0.59375, "learning_rate": 7.24804593733619e-05, "loss": 2.6125, "step": 2470 }, { "epoch": 0.31, "grad_norm": 0.6171875, "learning_rate": 7.233671201646985e-05, "loss": 2.6625, "step": 2480 }, { "epoch": 0.31125, "grad_norm": 0.6484375, "learning_rate": 7.21925211977624e-05, "loss": 2.6125, "step": 2490 }, { "epoch": 0.3125, "grad_norm": 0.6328125, "learning_rate": 7.20478892563329e-05, "loss": 2.6375, "step": 2500 }, { "epoch": 0.31375, "grad_norm": 0.6171875, "learning_rate": 7.190281853843075e-05, "loss": 2.6375, "step": 2510 }, { "epoch": 0.315, "grad_norm": 0.6484375, "learning_rate": 7.175731139742324e-05, "loss": 2.6625, "step": 2520 }, { "epoch": 0.31625, "grad_norm": 0.6640625, "learning_rate": 7.16113701937574e-05, "loss": 2.65, "step": 2530 }, { "epoch": 0.3175, "grad_norm": 0.67578125, "learning_rate": 7.146499729492174e-05, "loss": 2.6125, "step": 2540 }, { "epoch": 0.31875, "grad_norm": 0.671875, "learning_rate": 7.131819507540777e-05, "loss": 2.65, "step": 2550 }, { "epoch": 0.32, "grad_norm": 0.59375, "learning_rate": 7.117096591667157e-05, "loss": 2.625, "step": 2560 }, { "epoch": 0.32125, "grad_norm": 0.625, "learning_rate": 7.102331220709512e-05, "loss": 2.65, "step": 2570 }, { "epoch": 0.3225, "grad_norm": 0.671875, "learning_rate": 7.087523634194755e-05, "loss": 2.6375, "step": 2580 }, { "epoch": 0.32375, "grad_norm": 0.703125, "learning_rate": 7.072674072334624e-05, "loss": 2.6875, "step": 2590 }, { "epoch": 0.325, "grad_norm": 0.6484375, "learning_rate": 7.057782776021794e-05, "loss": 2.6375, "step": 2600 }, { "epoch": 0.32625, "grad_norm": 0.578125, "learning_rate": 7.042849986825964e-05, "loss": 2.7, "step": 2610 }, { "epoch": 0.3275, "grad_norm": 0.58203125, "learning_rate": 7.027875946989943e-05, "loss": 2.6375, "step": 2620 }, { "epoch": 0.32875, "grad_norm": 0.6328125, "learning_rate": 7.012860899425707e-05, "loss": 2.6625, "step": 2630 }, { "epoch": 0.33, "grad_norm": 0.6796875, "learning_rate": 6.997805087710478e-05, "loss": 2.6125, "step": 2640 }, { "epoch": 0.33125, "grad_norm": 0.62890625, "learning_rate": 6.982708756082756e-05, "loss": 2.6375, "step": 2650 }, { "epoch": 0.3325, "grad_norm": 0.68359375, "learning_rate": 6.967572149438366e-05, "loss": 2.65, "step": 2660 }, { "epoch": 0.33375, "grad_norm": 0.7109375, "learning_rate": 6.952395513326483e-05, "loss": 2.65, "step": 2670 }, { "epoch": 0.335, "grad_norm": 0.671875, "learning_rate": 6.937179093945646e-05, "loss": 2.625, "step": 2680 }, { "epoch": 0.33625, "grad_norm": 0.609375, "learning_rate": 6.92192313813977e-05, "loss": 2.65, "step": 2690 }, { "epoch": 0.3375, "grad_norm": 0.66015625, "learning_rate": 6.906627893394134e-05, "loss": 2.625, "step": 2700 }, { "epoch": 0.33875, "grad_norm": 0.625, "learning_rate": 6.891293607831373e-05, "loss": 2.675, "step": 2710 }, { "epoch": 0.34, "grad_norm": 0.6484375, "learning_rate": 6.875920530207452e-05, "loss": 2.6625, "step": 2720 }, { "epoch": 0.34125, "grad_norm": 0.6015625, "learning_rate": 6.860508909907623e-05, "loss": 2.6375, "step": 2730 }, { "epoch": 0.3425, "grad_norm": 0.65625, "learning_rate": 6.845058996942391e-05, "loss": 2.6, "step": 2740 }, { "epoch": 0.34375, "grad_norm": 0.609375, "learning_rate": 6.829571041943447e-05, "loss": 2.65, "step": 2750 }, { "epoch": 0.345, "grad_norm": 0.6875, "learning_rate": 6.814045296159618e-05, "loss": 2.675, "step": 2760 }, { "epoch": 0.34625, "grad_norm": 0.625, "learning_rate": 6.798482011452769e-05, "loss": 2.65, "step": 2770 }, { "epoch": 0.3475, "grad_norm": 0.62890625, "learning_rate": 6.782881440293734e-05, "loss": 2.625, "step": 2780 }, { "epoch": 0.34875, "grad_norm": 0.6484375, "learning_rate": 6.767243835758218e-05, "loss": 2.625, "step": 2790 }, { "epoch": 0.35, "grad_norm": 0.6875, "learning_rate": 6.751569451522686e-05, "loss": 2.6875, "step": 2800 }, { "epoch": 0.35125, "grad_norm": 0.640625, "learning_rate": 6.735858541860253e-05, "loss": 2.65, "step": 2810 }, { "epoch": 0.3525, "grad_norm": 0.6640625, "learning_rate": 6.720111361636554e-05, "loss": 2.6375, "step": 2820 }, { "epoch": 0.35375, "grad_norm": 0.6328125, "learning_rate": 6.704328166305616e-05, "loss": 2.6, "step": 2830 }, { "epoch": 0.355, "grad_norm": 0.63671875, "learning_rate": 6.688509211905707e-05, "loss": 2.675, "step": 2840 }, { "epoch": 0.35625, "grad_norm": 0.6328125, "learning_rate": 6.672654755055192e-05, "loss": 2.625, "step": 2850 }, { "epoch": 0.3575, "grad_norm": 0.578125, "learning_rate": 6.656765052948358e-05, "loss": 2.6, "step": 2860 }, { "epoch": 0.35875, "grad_norm": 0.609375, "learning_rate": 6.64084036335125e-05, "loss": 2.65, "step": 2870 }, { "epoch": 0.36, "grad_norm": 0.6328125, "learning_rate": 6.624880944597487e-05, "loss": 2.675, "step": 2880 }, { "epoch": 0.36125, "grad_norm": 0.6484375, "learning_rate": 6.608887055584072e-05, "loss": 2.65, "step": 2890 }, { "epoch": 0.3625, "grad_norm": 0.609375, "learning_rate": 6.592858955767192e-05, "loss": 2.6375, "step": 2900 }, { "epoch": 0.36375, "grad_norm": 0.66015625, "learning_rate": 6.57679690515801e-05, "loss": 2.6625, "step": 2910 }, { "epoch": 0.365, "grad_norm": 0.63671875, "learning_rate": 6.560701164318443e-05, "loss": 2.7125, "step": 2920 }, { "epoch": 0.36625, "grad_norm": 0.71484375, "learning_rate": 6.54457199435694e-05, "loss": 2.6125, "step": 2930 }, { "epoch": 0.3675, "grad_norm": 0.6875, "learning_rate": 6.528409656924242e-05, "loss": 2.6375, "step": 2940 }, { "epoch": 0.36875, "grad_norm": 0.58984375, "learning_rate": 6.51221441420914e-05, "loss": 2.6875, "step": 2950 }, { "epoch": 0.37, "grad_norm": 0.6484375, "learning_rate": 6.495986528934225e-05, "loss": 2.675, "step": 2960 }, { "epoch": 0.37125, "grad_norm": 0.578125, "learning_rate": 6.479726264351619e-05, "loss": 2.6375, "step": 2970 }, { "epoch": 0.3725, "grad_norm": 0.671875, "learning_rate": 6.463433884238708e-05, "loss": 2.625, "step": 2980 }, { "epoch": 0.37375, "grad_norm": 0.6484375, "learning_rate": 6.447109652893862e-05, "loss": 2.65, "step": 2990 }, { "epoch": 0.375, "grad_norm": 0.625, "learning_rate": 6.43075383513215e-05, "loss": 2.65, "step": 3000 }, { "epoch": 0.37625, "grad_norm": 0.5859375, "learning_rate": 6.414366696281045e-05, "loss": 2.675, "step": 3010 }, { "epoch": 0.3775, "grad_norm": 0.64453125, "learning_rate": 6.397948502176111e-05, "loss": 2.6625, "step": 3020 }, { "epoch": 0.37875, "grad_norm": 0.6328125, "learning_rate": 6.381499519156701e-05, "loss": 2.6625, "step": 3030 }, { "epoch": 0.38, "grad_norm": 0.6953125, "learning_rate": 6.365020014061632e-05, "loss": 2.675, "step": 3040 }, { "epoch": 0.38125, "grad_norm": 0.62109375, "learning_rate": 6.348510254224853e-05, "loss": 2.725, "step": 3050 }, { "epoch": 0.3825, "grad_norm": 0.625, "learning_rate": 6.331970507471116e-05, "loss": 2.675, "step": 3060 }, { "epoch": 0.38375, "grad_norm": 0.625, "learning_rate": 6.315401042111627e-05, "loss": 2.6625, "step": 3070 }, { "epoch": 0.385, "grad_norm": 0.63671875, "learning_rate": 6.298802126939689e-05, "loss": 2.6375, "step": 3080 }, { "epoch": 0.38625, "grad_norm": 0.625, "learning_rate": 6.282174031226349e-05, "loss": 2.5875, "step": 3090 }, { "epoch": 0.3875, "grad_norm": 0.6171875, "learning_rate": 6.265517024716027e-05, "loss": 2.6125, "step": 3100 }, { "epoch": 0.38875, "grad_norm": 0.65234375, "learning_rate": 6.24883137762214e-05, "loss": 2.6375, "step": 3110 }, { "epoch": 0.39, "grad_norm": 0.68359375, "learning_rate": 6.232117360622716e-05, "loss": 2.5875, "step": 3120 }, { "epoch": 0.39125, "grad_norm": 0.6484375, "learning_rate": 6.21537524485601e-05, "loss": 2.625, "step": 3130 }, { "epoch": 0.3925, "grad_norm": 0.66015625, "learning_rate": 6.198605301916095e-05, "loss": 2.6375, "step": 3140 }, { "epoch": 0.39375, "grad_norm": 0.62109375, "learning_rate": 6.181807803848469e-05, "loss": 2.6625, "step": 3150 }, { "epoch": 0.395, "grad_norm": 0.609375, "learning_rate": 6.16498302314563e-05, "loss": 2.75, "step": 3160 }, { "epoch": 0.39625, "grad_norm": 0.64453125, "learning_rate": 6.14813123274266e-05, "loss": 2.675, "step": 3170 }, { "epoch": 0.3975, "grad_norm": 0.6875, "learning_rate": 6.131252706012806e-05, "loss": 2.7125, "step": 3180 }, { "epoch": 0.39875, "grad_norm": 0.6484375, "learning_rate": 6.114347716763026e-05, "loss": 2.65, "step": 3190 }, { "epoch": 0.4, "grad_norm": 0.59765625, "learning_rate": 6.097416539229563e-05, "loss": 2.675, "step": 3200 }, { "epoch": 0.40125, "grad_norm": 0.6015625, "learning_rate": 6.080459448073498e-05, "loss": 2.5875, "step": 3210 }, { "epoch": 0.4025, "grad_norm": 0.67578125, "learning_rate": 6.06347671837628e-05, "loss": 2.65, "step": 3220 }, { "epoch": 0.40375, "grad_norm": 0.69140625, "learning_rate": 6.046468625635274e-05, "loss": 2.675, "step": 3230 }, { "epoch": 0.405, "grad_norm": 0.671875, "learning_rate": 6.0294354457592884e-05, "loss": 2.7125, "step": 3240 }, { "epoch": 0.40625, "grad_norm": 0.66796875, "learning_rate": 6.012377455064105e-05, "loss": 2.6375, "step": 3250 }, { "epoch": 0.4075, "grad_norm": 0.640625, "learning_rate": 5.995294930267988e-05, "loss": 2.575, "step": 3260 }, { "epoch": 0.40875, "grad_norm": 0.65234375, "learning_rate": 5.978188148487201e-05, "loss": 2.65, "step": 3270 }, { "epoch": 0.41, "grad_norm": 0.62109375, "learning_rate": 5.961057387231505e-05, "loss": 2.625, "step": 3280 }, { "epoch": 0.41125, "grad_norm": 0.64453125, "learning_rate": 5.9439029243996674e-05, "loss": 2.6875, "step": 3290 }, { "epoch": 0.4125, "grad_norm": 0.625, "learning_rate": 5.9267250382749405e-05, "loss": 2.675, "step": 3300 }, { "epoch": 0.41375, "grad_norm": 0.60546875, "learning_rate": 5.9095240075205615e-05, "loss": 2.6625, "step": 3310 }, { "epoch": 0.415, "grad_norm": 0.6484375, "learning_rate": 5.89230011117522e-05, "loss": 2.675, "step": 3320 }, { "epoch": 0.41625, "grad_norm": 0.625, "learning_rate": 5.875053628648535e-05, "loss": 2.6375, "step": 3330 }, { "epoch": 0.4175, "grad_norm": 0.59765625, "learning_rate": 5.8577848397165283e-05, "loss": 2.625, "step": 3340 }, { "epoch": 0.41875, "grad_norm": 0.6875, "learning_rate": 5.840494024517075e-05, "loss": 2.6375, "step": 3350 }, { "epoch": 0.42, "grad_norm": 0.59765625, "learning_rate": 5.823181463545368e-05, "loss": 2.6125, "step": 3360 }, { "epoch": 0.42125, "grad_norm": 0.640625, "learning_rate": 5.805847437649365e-05, "loss": 2.6875, "step": 3370 }, { "epoch": 0.4225, "grad_norm": 0.640625, "learning_rate": 5.788492228025233e-05, "loss": 2.6375, "step": 3380 }, { "epoch": 0.42375, "grad_norm": 0.65234375, "learning_rate": 5.771116116212781e-05, "loss": 2.625, "step": 3390 }, { "epoch": 0.425, "grad_norm": 0.6328125, "learning_rate": 5.7537193840909004e-05, "loss": 2.6375, "step": 3400 }, { "epoch": 0.42625, "grad_norm": 0.64453125, "learning_rate": 5.736302313872993e-05, "loss": 2.625, "step": 3410 }, { "epoch": 0.4275, "grad_norm": 0.609375, "learning_rate": 5.7188651881023854e-05, "loss": 2.5875, "step": 3420 }, { "epoch": 0.42875, "grad_norm": 0.67578125, "learning_rate": 5.7014082896477504e-05, "loss": 2.6125, "step": 3430 }, { "epoch": 0.43, "grad_norm": 0.59375, "learning_rate": 5.683931901698517e-05, "loss": 2.6875, "step": 3440 }, { "epoch": 0.43125, "grad_norm": 0.6328125, "learning_rate": 5.666436307760277e-05, "loss": 2.6375, "step": 3450 }, { "epoch": 0.4325, "grad_norm": 0.62890625, "learning_rate": 5.6489217916501874e-05, "loss": 2.625, "step": 3460 }, { "epoch": 0.43375, "grad_norm": 0.703125, "learning_rate": 5.6313886374923614e-05, "loss": 2.6375, "step": 3470 }, { "epoch": 0.435, "grad_norm": 0.65234375, "learning_rate": 5.613837129713266e-05, "loss": 2.6375, "step": 3480 }, { "epoch": 0.43625, "grad_norm": 0.6015625, "learning_rate": 5.596267553037102e-05, "loss": 2.675, "step": 3490 }, { "epoch": 0.4375, "grad_norm": 0.640625, "learning_rate": 5.578680192481188e-05, "loss": 2.675, "step": 3500 }, { "epoch": 0.43875, "grad_norm": 0.671875, "learning_rate": 5.561075333351338e-05, "loss": 2.625, "step": 3510 }, { "epoch": 0.44, "grad_norm": 0.671875, "learning_rate": 5.54345326123723e-05, "loss": 2.675, "step": 3520 }, { "epoch": 0.44125, "grad_norm": 0.66015625, "learning_rate": 5.525814262007775e-05, "loss": 2.6, "step": 3530 }, { "epoch": 0.4425, "grad_norm": 0.625, "learning_rate": 5.508158621806477e-05, "loss": 2.65, "step": 3540 }, { "epoch": 0.44375, "grad_norm": 0.5859375, "learning_rate": 5.4904866270467964e-05, "loss": 2.6875, "step": 3550 }, { "epoch": 0.445, "grad_norm": 0.6171875, "learning_rate": 5.472798564407502e-05, "loss": 2.6, "step": 3560 }, { "epoch": 0.44625, "grad_norm": 0.6484375, "learning_rate": 5.455094720828014e-05, "loss": 2.6, "step": 3570 }, { "epoch": 0.4475, "grad_norm": 0.58984375, "learning_rate": 5.437375383503758e-05, "loss": 2.7125, "step": 3580 }, { "epoch": 0.44875, "grad_norm": 0.609375, "learning_rate": 5.419640839881499e-05, "loss": 2.6, "step": 3590 }, { "epoch": 0.45, "grad_norm": 0.64453125, "learning_rate": 5.401891377654689e-05, "loss": 2.625, "step": 3600 }, { "epoch": 0.45125, "grad_norm": 0.640625, "learning_rate": 5.384127284758782e-05, "loss": 2.6, "step": 3610 }, { "epoch": 0.4525, "grad_norm": 0.67578125, "learning_rate": 5.366348849366584e-05, "loss": 2.6875, "step": 3620 }, { "epoch": 0.45375, "grad_norm": 0.59765625, "learning_rate": 5.348556359883562e-05, "loss": 2.575, "step": 3630 }, { "epoch": 0.455, "grad_norm": 0.6171875, "learning_rate": 5.330750104943171e-05, "loss": 2.6625, "step": 3640 }, { "epoch": 0.45625, "grad_norm": 0.6328125, "learning_rate": 5.312930373402175e-05, "loss": 2.65, "step": 3650 }, { "epoch": 0.4575, "grad_norm": 0.6484375, "learning_rate": 5.2950974543359585e-05, "loss": 2.625, "step": 3660 }, { "epoch": 0.45875, "grad_norm": 0.62109375, "learning_rate": 5.277251637033835e-05, "loss": 2.6, "step": 3670 }, { "epoch": 0.46, "grad_norm": 0.640625, "learning_rate": 5.259393210994355e-05, "loss": 2.5875, "step": 3680 }, { "epoch": 0.46125, "grad_norm": 0.625, "learning_rate": 5.241522465920611e-05, "loss": 2.6875, "step": 3690 }, { "epoch": 0.4625, "grad_norm": 0.625, "learning_rate": 5.2236396917155406e-05, "loss": 2.6125, "step": 3700 }, { "epoch": 0.46375, "grad_norm": 0.60546875, "learning_rate": 5.205745178477218e-05, "loss": 2.625, "step": 3710 }, { "epoch": 0.465, "grad_norm": 0.609375, "learning_rate": 5.187839216494149e-05, "loss": 2.6125, "step": 3720 }, { "epoch": 0.46625, "grad_norm": 0.64453125, "learning_rate": 5.169922096240565e-05, "loss": 2.6125, "step": 3730 }, { "epoch": 0.4675, "grad_norm": 0.609375, "learning_rate": 5.15199410837171e-05, "loss": 2.65, "step": 3740 }, { "epoch": 0.46875, "grad_norm": 0.65234375, "learning_rate": 5.134055543719122e-05, "loss": 2.6875, "step": 3750 }, { "epoch": 0.47, "grad_norm": 0.6640625, "learning_rate": 5.1161066932859226e-05, "loss": 2.6375, "step": 3760 }, { "epoch": 0.47125, "grad_norm": 0.63671875, "learning_rate": 5.098147848242085e-05, "loss": 2.575, "step": 3770 }, { "epoch": 0.4725, "grad_norm": 0.59375, "learning_rate": 5.0801792999197236e-05, "loss": 2.6625, "step": 3780 }, { "epoch": 0.47375, "grad_norm": 0.5859375, "learning_rate": 5.062201339808356e-05, "loss": 2.65, "step": 3790 }, { "epoch": 0.475, "grad_norm": 0.6796875, "learning_rate": 5.044214259550181e-05, "loss": 2.6375, "step": 3800 }, { "epoch": 0.47625, "grad_norm": 0.65234375, "learning_rate": 5.0262183509353507e-05, "loss": 2.65, "step": 3810 }, { "epoch": 0.4775, "grad_norm": 0.6015625, "learning_rate": 5.008213905897227e-05, "loss": 2.625, "step": 3820 }, { "epoch": 0.47875, "grad_norm": 0.68359375, "learning_rate": 4.990201216507655e-05, "loss": 2.6375, "step": 3830 }, { "epoch": 0.48, "grad_norm": 0.6328125, "learning_rate": 4.972180574972218e-05, "loss": 2.6625, "step": 3840 }, { "epoch": 0.48125, "grad_norm": 0.64453125, "learning_rate": 4.954152273625504e-05, "loss": 2.6, "step": 3850 }, { "epoch": 0.4825, "grad_norm": 0.671875, "learning_rate": 4.93611660492636e-05, "loss": 2.5875, "step": 3860 }, { "epoch": 0.48375, "grad_norm": 0.6328125, "learning_rate": 4.918073861453145e-05, "loss": 2.6375, "step": 3870 }, { "epoch": 0.485, "grad_norm": 0.6015625, "learning_rate": 4.900024335898988e-05, "loss": 2.6125, "step": 3880 }, { "epoch": 0.48625, "grad_norm": 0.6640625, "learning_rate": 4.881968321067038e-05, "loss": 2.625, "step": 3890 }, { "epoch": 0.4875, "grad_norm": 0.609375, "learning_rate": 4.863906109865715e-05, "loss": 2.6, "step": 3900 }, { "epoch": 0.48875, "grad_norm": 0.67578125, "learning_rate": 4.8458379953039594e-05, "loss": 2.6375, "step": 3910 }, { "epoch": 0.49, "grad_norm": 0.67578125, "learning_rate": 4.827764270486473e-05, "loss": 2.65, "step": 3920 }, { "epoch": 0.49125, "grad_norm": 0.59765625, "learning_rate": 4.8096852286089714e-05, "loss": 2.5625, "step": 3930 }, { "epoch": 0.4925, "grad_norm": 0.6171875, "learning_rate": 4.7916011629534216e-05, "loss": 2.625, "step": 3940 }, { "epoch": 0.49375, "grad_norm": 0.65625, "learning_rate": 4.773512366883294e-05, "loss": 2.6375, "step": 3950 }, { "epoch": 0.495, "grad_norm": 0.70703125, "learning_rate": 4.755419133838789e-05, "loss": 2.625, "step": 3960 }, { "epoch": 0.49625, "grad_norm": 0.6171875, "learning_rate": 4.737321757332088e-05, "loss": 2.625, "step": 3970 }, { "epoch": 0.4975, "grad_norm": 0.59375, "learning_rate": 4.7192205309425905e-05, "loss": 2.6375, "step": 3980 }, { "epoch": 0.49875, "grad_norm": 0.6171875, "learning_rate": 4.7011157483121454e-05, "loss": 2.625, "step": 3990 }, { "epoch": 0.5, "grad_norm": 0.68359375, "learning_rate": 4.6830077031402935e-05, "loss": 2.675, "step": 4000 }, { "epoch": 0.50125, "grad_norm": 0.6171875, "learning_rate": 4.6648966891795034e-05, "loss": 2.65, "step": 4010 }, { "epoch": 0.5025, "grad_norm": 0.69921875, "learning_rate": 4.6467830002304016e-05, "loss": 2.6375, "step": 4020 }, { "epoch": 0.50375, "grad_norm": 0.625, "learning_rate": 4.628666930137008e-05, "loss": 2.5875, "step": 4030 }, { "epoch": 0.505, "grad_norm": 0.58984375, "learning_rate": 4.610548772781972e-05, "loss": 2.65, "step": 4040 }, { "epoch": 0.50625, "grad_norm": 0.6171875, "learning_rate": 4.5924288220818024e-05, "loss": 2.625, "step": 4050 }, { "epoch": 0.5075, "grad_norm": 0.6328125, "learning_rate": 4.574307371982099e-05, "loss": 2.625, "step": 4060 }, { "epoch": 0.50875, "grad_norm": 0.6875, "learning_rate": 4.556184716452788e-05, "loss": 2.6, "step": 4070 }, { "epoch": 0.51, "grad_norm": 0.6328125, "learning_rate": 4.5380611494833464e-05, "loss": 2.675, "step": 4080 }, { "epoch": 0.51125, "grad_norm": 0.703125, "learning_rate": 4.5199369650780375e-05, "loss": 2.6, "step": 4090 }, { "epoch": 0.5125, "grad_norm": 0.59375, "learning_rate": 4.501812457251145e-05, "loss": 2.65, "step": 4100 }, { "epoch": 0.51375, "grad_norm": 0.64453125, "learning_rate": 4.4836879200221946e-05, "loss": 2.6, "step": 4110 }, { "epoch": 0.515, "grad_norm": 0.63671875, "learning_rate": 4.465563647411191e-05, "loss": 2.675, "step": 4120 }, { "epoch": 0.51625, "grad_norm": 0.6640625, "learning_rate": 4.447439933433846e-05, "loss": 2.6375, "step": 4130 }, { "epoch": 0.5175, "grad_norm": 0.625, "learning_rate": 4.429317072096808e-05, "loss": 2.5875, "step": 4140 }, { "epoch": 0.51875, "grad_norm": 0.625, "learning_rate": 4.4111953573928944e-05, "loss": 2.65, "step": 4150 }, { "epoch": 0.52, "grad_norm": 0.64453125, "learning_rate": 4.393075083296323e-05, "loss": 2.625, "step": 4160 }, { "epoch": 0.52125, "grad_norm": 0.6640625, "learning_rate": 4.3749565437579406e-05, "loss": 2.6, "step": 4170 }, { "epoch": 0.5225, "grad_norm": 0.6484375, "learning_rate": 4.3568400327004545e-05, "loss": 2.65, "step": 4180 }, { "epoch": 0.52375, "grad_norm": 0.6640625, "learning_rate": 4.3387258440136665e-05, "loss": 2.6125, "step": 4190 }, { "epoch": 0.525, "grad_norm": 0.60546875, "learning_rate": 4.320614271549707e-05, "loss": 2.5625, "step": 4200 }, { "epoch": 0.52625, "grad_norm": 0.625, "learning_rate": 4.302505609118261e-05, "loss": 2.6, "step": 4210 }, { "epoch": 0.5275, "grad_norm": 0.63671875, "learning_rate": 4.28440015048181e-05, "loss": 2.625, "step": 4220 }, { "epoch": 0.52875, "grad_norm": 0.6484375, "learning_rate": 4.26629818935086e-05, "loss": 2.6, "step": 4230 }, { "epoch": 0.53, "grad_norm": 0.65234375, "learning_rate": 4.24820001937918e-05, "loss": 2.6, "step": 4240 }, { "epoch": 0.53125, "grad_norm": 0.609375, "learning_rate": 4.230105934159042e-05, "loss": 2.6375, "step": 4250 }, { "epoch": 0.5325, "grad_norm": 0.6171875, "learning_rate": 4.2120162272164506e-05, "loss": 2.625, "step": 4260 }, { "epoch": 0.53375, "grad_norm": 0.62109375, "learning_rate": 4.1939311920063854e-05, "loss": 2.6125, "step": 4270 }, { "epoch": 0.535, "grad_norm": 0.625, "learning_rate": 4.175851121908038e-05, "loss": 2.5875, "step": 4280 }, { "epoch": 0.53625, "grad_norm": 0.609375, "learning_rate": 4.157776310220063e-05, "loss": 2.5875, "step": 4290 }, { "epoch": 0.5375, "grad_norm": 0.66015625, "learning_rate": 4.139707050155802e-05, "loss": 2.6, "step": 4300 }, { "epoch": 0.53875, "grad_norm": 0.65625, "learning_rate": 4.121643634838546e-05, "loss": 2.625, "step": 4310 }, { "epoch": 0.54, "grad_norm": 0.67578125, "learning_rate": 4.1035863572967634e-05, "loss": 2.6125, "step": 4320 }, { "epoch": 0.54125, "grad_norm": 0.62890625, "learning_rate": 4.085535510459359e-05, "loss": 2.625, "step": 4330 }, { "epoch": 0.5425, "grad_norm": 0.6328125, "learning_rate": 4.0674913871509165e-05, "loss": 2.575, "step": 4340 }, { "epoch": 0.54375, "grad_norm": 0.62109375, "learning_rate": 4.0494542800869474e-05, "loss": 2.65, "step": 4350 }, { "epoch": 0.545, "grad_norm": 0.59375, "learning_rate": 4.031424481869147e-05, "loss": 2.6, "step": 4360 }, { "epoch": 0.54625, "grad_norm": 0.65234375, "learning_rate": 4.0134022849806444e-05, "loss": 2.6, "step": 4370 }, { "epoch": 0.5475, "grad_norm": 0.6171875, "learning_rate": 3.995387981781256e-05, "loss": 2.65, "step": 4380 }, { "epoch": 0.54875, "grad_norm": 0.609375, "learning_rate": 3.9773818645027466e-05, "loss": 2.6375, "step": 4390 }, { "epoch": 0.55, "grad_norm": 0.6171875, "learning_rate": 3.9593842252440886e-05, "loss": 2.5625, "step": 4400 }, { "epoch": 0.55125, "grad_norm": 0.6328125, "learning_rate": 3.9413953559667205e-05, "loss": 2.6375, "step": 4410 }, { "epoch": 0.5525, "grad_norm": 0.609375, "learning_rate": 3.9234155484898134e-05, "loss": 2.625, "step": 4420 }, { "epoch": 0.55375, "grad_norm": 0.60546875, "learning_rate": 3.9054450944855346e-05, "loss": 2.65, "step": 4430 }, { "epoch": 0.555, "grad_norm": 0.6640625, "learning_rate": 3.8874842854743174e-05, "loss": 2.575, "step": 4440 }, { "epoch": 0.55625, "grad_norm": 0.640625, "learning_rate": 3.869533412820136e-05, "loss": 2.6, "step": 4450 }, { "epoch": 0.5575, "grad_norm": 0.61328125, "learning_rate": 3.85159276772577e-05, "loss": 2.575, "step": 4460 }, { "epoch": 0.55875, "grad_norm": 0.64453125, "learning_rate": 3.8336626412280865e-05, "loss": 2.6125, "step": 4470 }, { "epoch": 0.56, "grad_norm": 0.62890625, "learning_rate": 3.815743324193321e-05, "loss": 2.625, "step": 4480 }, { "epoch": 0.56125, "grad_norm": 0.609375, "learning_rate": 3.797835107312349e-05, "loss": 2.55, "step": 4490 }, { "epoch": 0.5625, "grad_norm": 0.6875, "learning_rate": 3.7799382810959854e-05, "loss": 2.625, "step": 4500 }, { "epoch": 0.56375, "grad_norm": 0.6328125, "learning_rate": 3.762053135870257e-05, "loss": 2.6375, "step": 4510 }, { "epoch": 0.565, "grad_norm": 0.62890625, "learning_rate": 3.7441799617717e-05, "loss": 2.625, "step": 4520 }, { "epoch": 0.56625, "grad_norm": 0.63671875, "learning_rate": 3.7263190487426565e-05, "loss": 2.6375, "step": 4530 }, { "epoch": 0.5675, "grad_norm": 0.6328125, "learning_rate": 3.708470686526565e-05, "loss": 2.525, "step": 4540 }, { "epoch": 0.56875, "grad_norm": 0.65625, "learning_rate": 3.690635164663257e-05, "loss": 2.7, "step": 4550 }, { "epoch": 0.57, "grad_norm": 0.609375, "learning_rate": 3.672812772484274e-05, "loss": 2.625, "step": 4560 }, { "epoch": 0.57125, "grad_norm": 0.74609375, "learning_rate": 3.655003799108159e-05, "loss": 2.6, "step": 4570 }, { "epoch": 0.5725, "grad_norm": 0.7578125, "learning_rate": 3.637208533435773e-05, "loss": 2.6125, "step": 4580 }, { "epoch": 0.57375, "grad_norm": 0.59765625, "learning_rate": 3.6194272641456096e-05, "loss": 2.575, "step": 4590 }, { "epoch": 0.575, "grad_norm": 0.671875, "learning_rate": 3.601660279689107e-05, "loss": 2.65, "step": 4600 }, { "epoch": 0.57625, "grad_norm": 0.64453125, "learning_rate": 3.583907868285979e-05, "loss": 2.6375, "step": 4610 }, { "epoch": 0.5775, "grad_norm": 0.6171875, "learning_rate": 3.566170317919523e-05, "loss": 2.65, "step": 4620 }, { "epoch": 0.57875, "grad_norm": 0.625, "learning_rate": 3.5484479163319646e-05, "loss": 2.5625, "step": 4630 }, { "epoch": 0.58, "grad_norm": 0.609375, "learning_rate": 3.5307409510197785e-05, "loss": 2.6125, "step": 4640 }, { "epoch": 0.58125, "grad_norm": 0.63671875, "learning_rate": 3.513049709229031e-05, "loss": 2.6375, "step": 4650 }, { "epoch": 0.5825, "grad_norm": 0.64453125, "learning_rate": 3.4953744779507194e-05, "loss": 2.6125, "step": 4660 }, { "epoch": 0.58375, "grad_norm": 0.6171875, "learning_rate": 3.4777155439161125e-05, "loss": 2.5875, "step": 4670 }, { "epoch": 0.585, "grad_norm": 0.578125, "learning_rate": 3.4600731935921026e-05, "loss": 2.625, "step": 4680 }, { "epoch": 0.58625, "grad_norm": 0.6171875, "learning_rate": 3.442447713176557e-05, "loss": 2.6625, "step": 4690 }, { "epoch": 0.5875, "grad_norm": 0.6328125, "learning_rate": 3.4248393885936794e-05, "loss": 2.5875, "step": 4700 }, { "epoch": 0.58875, "grad_norm": 0.703125, "learning_rate": 3.407248505489363e-05, "loss": 2.6625, "step": 4710 }, { "epoch": 0.59, "grad_norm": 0.67578125, "learning_rate": 3.389675349226566e-05, "loss": 2.625, "step": 4720 }, { "epoch": 0.59125, "grad_norm": 0.66015625, "learning_rate": 3.372120204880675e-05, "loss": 2.6875, "step": 4730 }, { "epoch": 0.5925, "grad_norm": 0.6484375, "learning_rate": 3.354583357234883e-05, "loss": 2.65, "step": 4740 }, { "epoch": 0.59375, "grad_norm": 0.6484375, "learning_rate": 3.3370650907755755e-05, "loss": 2.6625, "step": 4750 }, { "epoch": 0.595, "grad_norm": 0.6484375, "learning_rate": 3.3195656896877026e-05, "loss": 2.6, "step": 4760 }, { "epoch": 0.59625, "grad_norm": 0.58203125, "learning_rate": 3.3020854378501845e-05, "loss": 2.675, "step": 4770 }, { "epoch": 0.5975, "grad_norm": 0.671875, "learning_rate": 3.284624618831291e-05, "loss": 2.6, "step": 4780 }, { "epoch": 0.59875, "grad_norm": 0.625, "learning_rate": 3.267183515884054e-05, "loss": 2.6125, "step": 4790 }, { "epoch": 0.6, "grad_norm": 0.6796875, "learning_rate": 3.249762411941663e-05, "loss": 2.6125, "step": 4800 }, { "epoch": 0.60125, "grad_norm": 0.62109375, "learning_rate": 3.232361589612883e-05, "loss": 2.6, "step": 4810 }, { "epoch": 0.6025, "grad_norm": 0.63671875, "learning_rate": 3.2149813311774644e-05, "loss": 2.675, "step": 4820 }, { "epoch": 0.60375, "grad_norm": 0.64453125, "learning_rate": 3.1976219185815656e-05, "loss": 2.6125, "step": 4830 }, { "epoch": 0.605, "grad_norm": 0.59765625, "learning_rate": 3.180283633433182e-05, "loss": 2.6125, "step": 4840 }, { "epoch": 0.60625, "grad_norm": 0.59765625, "learning_rate": 3.16296675699757e-05, "loss": 2.5375, "step": 4850 }, { "epoch": 0.6075, "grad_norm": 0.64453125, "learning_rate": 3.1456715701926975e-05, "loss": 2.5625, "step": 4860 }, { "epoch": 0.60875, "grad_norm": 0.6328125, "learning_rate": 3.128398353584671e-05, "loss": 2.575, "step": 4870 }, { "epoch": 0.61, "grad_norm": 0.58203125, "learning_rate": 3.1111473873831967e-05, "loss": 2.6625, "step": 4880 }, { "epoch": 0.61125, "grad_norm": 0.703125, "learning_rate": 3.0939189514370264e-05, "loss": 2.5875, "step": 4890 }, { "epoch": 0.6125, "grad_norm": 0.66015625, "learning_rate": 3.076713325229424e-05, "loss": 2.6125, "step": 4900 }, { "epoch": 0.61375, "grad_norm": 0.65625, "learning_rate": 3.0595307878736294e-05, "loss": 2.625, "step": 4910 }, { "epoch": 0.615, "grad_norm": 0.66015625, "learning_rate": 3.0423716181083286e-05, "loss": 2.6, "step": 4920 }, { "epoch": 0.61625, "grad_norm": 0.63671875, "learning_rate": 3.025236094293135e-05, "loss": 2.6, "step": 4930 }, { "epoch": 0.6175, "grad_norm": 0.6796875, "learning_rate": 3.0081244944040706e-05, "loss": 2.5875, "step": 4940 }, { "epoch": 0.61875, "grad_norm": 0.6484375, "learning_rate": 2.991037096029062e-05, "loss": 2.6, "step": 4950 }, { "epoch": 0.62, "grad_norm": 0.64453125, "learning_rate": 2.9739741763634308e-05, "loss": 2.6375, "step": 4960 }, { "epoch": 0.62125, "grad_norm": 0.64453125, "learning_rate": 2.9569360122054008e-05, "loss": 2.5625, "step": 4970 }, { "epoch": 0.6225, "grad_norm": 0.65625, "learning_rate": 2.9399228799516064e-05, "loss": 2.6625, "step": 4980 }, { "epoch": 0.62375, "grad_norm": 0.6171875, "learning_rate": 2.92293505559261e-05, "loss": 2.55, "step": 4990 }, { "epoch": 0.625, "grad_norm": 0.6171875, "learning_rate": 2.9059728147084234e-05, "loss": 2.625, "step": 5000 }, { "epoch": 0.62625, "grad_norm": 0.64453125, "learning_rate": 2.8890364324640378e-05, "loss": 2.55, "step": 5010 }, { "epoch": 0.6275, "grad_norm": 0.609375, "learning_rate": 2.8721261836049633e-05, "loss": 2.5875, "step": 5020 }, { "epoch": 0.62875, "grad_norm": 0.62109375, "learning_rate": 2.8552423424527645e-05, "loss": 2.6125, "step": 5030 }, { "epoch": 0.63, "grad_norm": 0.6640625, "learning_rate": 2.838385182900617e-05, "loss": 2.6, "step": 5040 }, { "epoch": 0.63125, "grad_norm": 0.64453125, "learning_rate": 2.8215549784088596e-05, "loss": 2.6375, "step": 5050 }, { "epoch": 0.6325, "grad_norm": 0.62890625, "learning_rate": 2.804752002000564e-05, "loss": 2.575, "step": 5060 }, { "epoch": 0.63375, "grad_norm": 0.62890625, "learning_rate": 2.787976526257101e-05, "loss": 2.6625, "step": 5070 }, { "epoch": 0.635, "grad_norm": 0.625, "learning_rate": 2.771228823313718e-05, "loss": 2.625, "step": 5080 }, { "epoch": 0.63625, "grad_norm": 0.6484375, "learning_rate": 2.7545091648551278e-05, "loss": 2.575, "step": 5090 }, { "epoch": 0.6375, "grad_norm": 0.6171875, "learning_rate": 2.7378178221111e-05, "loss": 2.6, "step": 5100 }, { "epoch": 0.63875, "grad_norm": 0.66015625, "learning_rate": 2.721155065852061e-05, "loss": 2.5875, "step": 5110 }, { "epoch": 0.64, "grad_norm": 0.609375, "learning_rate": 2.704521166384701e-05, "loss": 2.6125, "step": 5120 }, { "epoch": 0.64125, "grad_norm": 0.62890625, "learning_rate": 2.6879163935475876e-05, "loss": 2.6875, "step": 5130 }, { "epoch": 0.6425, "grad_norm": 0.640625, "learning_rate": 2.6713410167067923e-05, "loss": 2.6, "step": 5140 }, { "epoch": 0.64375, "grad_norm": 0.609375, "learning_rate": 2.6547953047515193e-05, "loss": 2.5625, "step": 5150 }, { "epoch": 0.645, "grad_norm": 0.625, "learning_rate": 2.6382795260897407e-05, "loss": 2.5625, "step": 5160 }, { "epoch": 0.64625, "grad_norm": 0.59375, "learning_rate": 2.6217939486438454e-05, "loss": 2.575, "step": 5170 }, { "epoch": 0.6475, "grad_norm": 0.64453125, "learning_rate": 2.6053388398462907e-05, "loss": 2.65, "step": 5180 }, { "epoch": 0.64875, "grad_norm": 0.66015625, "learning_rate": 2.5889144666352694e-05, "loss": 2.625, "step": 5190 }, { "epoch": 0.65, "grad_norm": 0.66015625, "learning_rate": 2.5725210954503688e-05, "loss": 2.575, "step": 5200 }, { "epoch": 0.65125, "grad_norm": 0.59765625, "learning_rate": 2.5561589922282595e-05, "loss": 2.5875, "step": 5210 }, { "epoch": 0.6525, "grad_norm": 0.6875, "learning_rate": 2.5398284223983782e-05, "loss": 2.6375, "step": 5220 }, { "epoch": 0.65375, "grad_norm": 0.6328125, "learning_rate": 2.5235296508786133e-05, "loss": 2.6375, "step": 5230 }, { "epoch": 0.655, "grad_norm": 0.6171875, "learning_rate": 2.5072629420710227e-05, "loss": 2.55, "step": 5240 }, { "epoch": 0.65625, "grad_norm": 0.6328125, "learning_rate": 2.4910285598575273e-05, "loss": 2.5875, "step": 5250 }, { "epoch": 0.6575, "grad_norm": 0.6328125, "learning_rate": 2.474826767595651e-05, "loss": 2.6125, "step": 5260 }, { "epoch": 0.65875, "grad_norm": 0.6171875, "learning_rate": 2.4586578281142255e-05, "loss": 2.575, "step": 5270 }, { "epoch": 0.66, "grad_norm": 0.671875, "learning_rate": 2.442522003709144e-05, "loss": 2.6125, "step": 5280 }, { "epoch": 0.66125, "grad_norm": 0.6171875, "learning_rate": 2.4264195561391e-05, "loss": 2.6125, "step": 5290 }, { "epoch": 0.6625, "grad_norm": 0.65234375, "learning_rate": 2.410350746621337e-05, "loss": 2.5875, "step": 5300 }, { "epoch": 0.66375, "grad_norm": 0.65625, "learning_rate": 2.3943158358274205e-05, "loss": 2.6, "step": 5310 }, { "epoch": 0.665, "grad_norm": 0.6171875, "learning_rate": 2.378315083878998e-05, "loss": 2.6125, "step": 5320 }, { "epoch": 0.66625, "grad_norm": 0.5625, "learning_rate": 2.3623487503435895e-05, "loss": 2.5375, "step": 5330 }, { "epoch": 0.6675, "grad_norm": 0.6484375, "learning_rate": 2.346417094230373e-05, "loss": 2.65, "step": 5340 }, { "epoch": 0.66875, "grad_norm": 0.625, "learning_rate": 2.3305203739859762e-05, "loss": 2.5875, "step": 5350 }, { "epoch": 0.67, "grad_norm": 0.6328125, "learning_rate": 2.3146588474902984e-05, "loss": 2.5625, "step": 5360 }, { "epoch": 0.67125, "grad_norm": 0.6640625, "learning_rate": 2.29883277205231e-05, "loss": 2.6, "step": 5370 }, { "epoch": 0.6725, "grad_norm": 0.609375, "learning_rate": 2.283042404405893e-05, "loss": 2.6, "step": 5380 }, { "epoch": 0.67375, "grad_norm": 0.6484375, "learning_rate": 2.2672880007056706e-05, "loss": 2.5625, "step": 5390 }, { "epoch": 0.675, "grad_norm": 0.61328125, "learning_rate": 2.251569816522844e-05, "loss": 2.6375, "step": 5400 }, { "epoch": 0.67625, "grad_norm": 0.58203125, "learning_rate": 2.235888106841063e-05, "loss": 2.5625, "step": 5410 }, { "epoch": 0.6775, "grad_norm": 0.609375, "learning_rate": 2.220243126052274e-05, "loss": 2.6625, "step": 5420 }, { "epoch": 0.67875, "grad_norm": 0.63671875, "learning_rate": 2.2046351279526028e-05, "loss": 2.625, "step": 5430 }, { "epoch": 0.68, "grad_norm": 0.59375, "learning_rate": 2.189064365738236e-05, "loss": 2.575, "step": 5440 }, { "epoch": 0.68125, "grad_norm": 0.61328125, "learning_rate": 2.1735310920013058e-05, "loss": 2.575, "step": 5450 }, { "epoch": 0.6825, "grad_norm": 0.67578125, "learning_rate": 2.1580355587258037e-05, "loss": 2.5875, "step": 5460 }, { "epoch": 0.68375, "grad_norm": 0.61328125, "learning_rate": 2.1425780172834884e-05, "loss": 2.5625, "step": 5470 }, { "epoch": 0.685, "grad_norm": 0.59765625, "learning_rate": 2.1271587184298014e-05, "loss": 2.6, "step": 5480 }, { "epoch": 0.68625, "grad_norm": 0.6640625, "learning_rate": 2.111777912299812e-05, "loss": 2.575, "step": 5490 }, { "epoch": 0.6875, "grad_norm": 0.62109375, "learning_rate": 2.0964358484041447e-05, "loss": 2.55, "step": 5500 }, { "epoch": 0.68875, "grad_norm": 0.64453125, "learning_rate": 2.08113277562495e-05, "loss": 2.55, "step": 5510 }, { "epoch": 0.69, "grad_norm": 0.59765625, "learning_rate": 2.0658689422118448e-05, "loss": 2.5625, "step": 5520 }, { "epoch": 0.69125, "grad_norm": 0.578125, "learning_rate": 2.0506445957779053e-05, "loss": 2.625, "step": 5530 }, { "epoch": 0.6925, "grad_norm": 0.62109375, "learning_rate": 2.0354599832956395e-05, "loss": 2.6, "step": 5540 }, { "epoch": 0.69375, "grad_norm": 0.62890625, "learning_rate": 2.0203153510929794e-05, "loss": 2.575, "step": 5550 }, { "epoch": 0.695, "grad_norm": 0.6640625, "learning_rate": 2.0052109448492946e-05, "loss": 2.5875, "step": 5560 }, { "epoch": 0.69625, "grad_norm": 0.59765625, "learning_rate": 1.9901470095913946e-05, "loss": 2.65, "step": 5570 }, { "epoch": 0.6975, "grad_norm": 0.65234375, "learning_rate": 1.9751237896895645e-05, "loss": 2.5375, "step": 5580 }, { "epoch": 0.69875, "grad_norm": 0.6171875, "learning_rate": 1.960141528853598e-05, "loss": 2.575, "step": 5590 }, { "epoch": 0.7, "grad_norm": 0.61328125, "learning_rate": 1.9452004701288352e-05, "loss": 2.625, "step": 5600 }, { "epoch": 0.70125, "grad_norm": 0.5859375, "learning_rate": 1.930300855892235e-05, "loss": 2.5625, "step": 5610 }, { "epoch": 0.7025, "grad_norm": 0.65234375, "learning_rate": 1.9154429278484284e-05, "loss": 2.6375, "step": 5620 }, { "epoch": 0.70375, "grad_norm": 0.625, "learning_rate": 1.900626927025809e-05, "loss": 2.6, "step": 5630 }, { "epoch": 0.705, "grad_norm": 0.6484375, "learning_rate": 1.8858530937726193e-05, "loss": 2.625, "step": 5640 }, { "epoch": 0.70625, "grad_norm": 0.6328125, "learning_rate": 1.8711216677530438e-05, "loss": 2.65, "step": 5650 }, { "epoch": 0.7075, "grad_norm": 0.70703125, "learning_rate": 1.856432887943336e-05, "loss": 2.525, "step": 5660 }, { "epoch": 0.70875, "grad_norm": 0.66796875, "learning_rate": 1.8417869926279277e-05, "loss": 2.5875, "step": 5670 }, { "epoch": 0.71, "grad_norm": 0.61328125, "learning_rate": 1.8271842193955737e-05, "loss": 2.6375, "step": 5680 }, { "epoch": 0.71125, "grad_norm": 0.6328125, "learning_rate": 1.8126248051354934e-05, "loss": 2.5625, "step": 5690 }, { "epoch": 0.7125, "grad_norm": 0.7109375, "learning_rate": 1.7981089860335227e-05, "loss": 2.65, "step": 5700 }, { "epoch": 0.71375, "grad_norm": 0.7109375, "learning_rate": 1.7836369975682933e-05, "loss": 2.575, "step": 5710 }, { "epoch": 0.715, "grad_norm": 0.6484375, "learning_rate": 1.769209074507407e-05, "loss": 2.6125, "step": 5720 }, { "epoch": 0.71625, "grad_norm": 0.61328125, "learning_rate": 1.7548254509036224e-05, "loss": 2.625, "step": 5730 }, { "epoch": 0.7175, "grad_norm": 0.6953125, "learning_rate": 1.7404863600910688e-05, "loss": 2.625, "step": 5740 }, { "epoch": 0.71875, "grad_norm": 0.60546875, "learning_rate": 1.7261920346814526e-05, "loss": 2.6125, "step": 5750 }, { "epoch": 0.72, "grad_norm": 0.61328125, "learning_rate": 1.7119427065602886e-05, "loss": 2.6125, "step": 5760 }, { "epoch": 0.72125, "grad_norm": 0.625, "learning_rate": 1.6977386068831323e-05, "loss": 2.65, "step": 5770 }, { "epoch": 0.7225, "grad_norm": 0.6484375, "learning_rate": 1.683579966071837e-05, "loss": 2.575, "step": 5780 }, { "epoch": 0.72375, "grad_norm": 0.65234375, "learning_rate": 1.669467013810814e-05, "loss": 2.575, "step": 5790 }, { "epoch": 0.725, "grad_norm": 0.6484375, "learning_rate": 1.655399979043301e-05, "loss": 2.525, "step": 5800 }, { "epoch": 0.72625, "grad_norm": 0.6640625, "learning_rate": 1.641379089967658e-05, "loss": 2.625, "step": 5810 }, { "epoch": 0.7275, "grad_norm": 0.62890625, "learning_rate": 1.6274045740336544e-05, "loss": 2.575, "step": 5820 }, { "epoch": 0.72875, "grad_norm": 0.6484375, "learning_rate": 1.6134766579387894e-05, "loss": 2.6, "step": 5830 }, { "epoch": 0.73, "grad_norm": 0.62890625, "learning_rate": 1.5995955676246095e-05, "loss": 2.6375, "step": 5840 }, { "epoch": 0.73125, "grad_norm": 0.67578125, "learning_rate": 1.5857615282730392e-05, "loss": 2.6125, "step": 5850 }, { "epoch": 0.7325, "grad_norm": 0.62109375, "learning_rate": 1.5719747643027396e-05, "loss": 2.5875, "step": 5860 }, { "epoch": 0.73375, "grad_norm": 0.56640625, "learning_rate": 1.558235499365453e-05, "loss": 2.55, "step": 5870 }, { "epoch": 0.735, "grad_norm": 0.6328125, "learning_rate": 1.544543956342389e-05, "loss": 2.6375, "step": 5880 }, { "epoch": 0.73625, "grad_norm": 0.59765625, "learning_rate": 1.5309003573406005e-05, "loss": 2.5875, "step": 5890 }, { "epoch": 0.7375, "grad_norm": 0.61328125, "learning_rate": 1.5173049236893795e-05, "loss": 2.6, "step": 5900 }, { "epoch": 0.73875, "grad_norm": 0.6171875, "learning_rate": 1.5037578759366741e-05, "loss": 2.5875, "step": 5910 }, { "epoch": 0.74, "grad_norm": 0.6328125, "learning_rate": 1.4902594338455027e-05, "loss": 2.575, "step": 5920 }, { "epoch": 0.74125, "grad_norm": 0.65234375, "learning_rate": 1.4768098163903951e-05, "loss": 2.5875, "step": 5930 }, { "epoch": 0.7425, "grad_norm": 0.6328125, "learning_rate": 1.463409241753838e-05, "loss": 2.525, "step": 5940 }, { "epoch": 0.74375, "grad_norm": 0.640625, "learning_rate": 1.4500579273227315e-05, "loss": 2.6125, "step": 5950 }, { "epoch": 0.745, "grad_norm": 0.65625, "learning_rate": 1.43675608968487e-05, "loss": 2.525, "step": 5960 }, { "epoch": 0.74625, "grad_norm": 0.609375, "learning_rate": 1.4235039446254265e-05, "loss": 2.575, "step": 5970 }, { "epoch": 0.7475, "grad_norm": 0.64453125, "learning_rate": 1.4103017071234433e-05, "loss": 2.5375, "step": 5980 }, { "epoch": 0.74875, "grad_norm": 0.62109375, "learning_rate": 1.3971495913483623e-05, "loss": 2.6, "step": 5990 }, { "epoch": 0.75, "grad_norm": 0.6484375, "learning_rate": 1.384047810656532e-05, "loss": 2.625, "step": 6000 }, { "epoch": 0.75125, "grad_norm": 0.6328125, "learning_rate": 1.3709965775877599e-05, "loss": 2.525, "step": 6010 }, { "epoch": 0.7525, "grad_norm": 0.63671875, "learning_rate": 1.3579961038618541e-05, "loss": 2.625, "step": 6020 }, { "epoch": 0.75375, "grad_norm": 0.625, "learning_rate": 1.3450466003751991e-05, "loss": 2.6125, "step": 6030 }, { "epoch": 0.755, "grad_norm": 0.64453125, "learning_rate": 1.3321482771973297e-05, "loss": 2.6, "step": 6040 }, { "epoch": 0.75625, "grad_norm": 0.65234375, "learning_rate": 1.3193013435675177e-05, "loss": 2.6, "step": 6050 }, { "epoch": 0.7575, "grad_norm": 0.63671875, "learning_rate": 1.3065060078913894e-05, "loss": 2.6125, "step": 6060 }, { "epoch": 0.75875, "grad_norm": 0.6015625, "learning_rate": 1.2937624777375331e-05, "loss": 2.575, "step": 6070 }, { "epoch": 0.76, "grad_norm": 0.6171875, "learning_rate": 1.2810709598341405e-05, "loss": 2.5875, "step": 6080 }, { "epoch": 0.76125, "grad_norm": 0.609375, "learning_rate": 1.268431660065651e-05, "loss": 2.65, "step": 6090 }, { "epoch": 0.7625, "grad_norm": 0.58984375, "learning_rate": 1.2558447834694046e-05, "loss": 2.5125, "step": 6100 }, { "epoch": 0.76375, "grad_norm": 0.61328125, "learning_rate": 1.2433105342323292e-05, "loss": 2.65, "step": 6110 }, { "epoch": 0.765, "grad_norm": 0.69140625, "learning_rate": 1.2308291156876137e-05, "loss": 2.575, "step": 6120 }, { "epoch": 0.76625, "grad_norm": 0.60546875, "learning_rate": 1.2184007303114218e-05, "loss": 2.5625, "step": 6130 }, { "epoch": 0.7675, "grad_norm": 0.69921875, "learning_rate": 1.2060255797196016e-05, "loss": 2.6125, "step": 6140 }, { "epoch": 0.76875, "grad_norm": 0.671875, "learning_rate": 1.193703864664411e-05, "loss": 2.6375, "step": 6150 }, { "epoch": 0.77, "grad_norm": 0.609375, "learning_rate": 1.181435785031272e-05, "loss": 2.5875, "step": 6160 }, { "epoch": 0.77125, "grad_norm": 0.65234375, "learning_rate": 1.169221539835515e-05, "loss": 2.5375, "step": 6170 }, { "epoch": 0.7725, "grad_norm": 0.60546875, "learning_rate": 1.1570613272191628e-05, "loss": 2.625, "step": 6180 }, { "epoch": 0.77375, "grad_norm": 0.671875, "learning_rate": 1.1449553444477083e-05, "loss": 2.525, "step": 6190 }, { "epoch": 0.775, "grad_norm": 0.62109375, "learning_rate": 1.1329037879069135e-05, "loss": 2.575, "step": 6200 }, { "epoch": 0.77625, "grad_norm": 0.6015625, "learning_rate": 1.120906853099632e-05, "loss": 2.5375, "step": 6210 }, { "epoch": 0.7775, "grad_norm": 0.625, "learning_rate": 1.1089647346426304e-05, "loss": 2.575, "step": 6220 }, { "epoch": 0.77875, "grad_norm": 0.59765625, "learning_rate": 1.0970776262634293e-05, "loss": 2.6625, "step": 6230 }, { "epoch": 0.78, "grad_norm": 0.625, "learning_rate": 1.0852457207971724e-05, "loss": 2.6, "step": 6240 }, { "epoch": 0.78125, "grad_norm": 0.65234375, "learning_rate": 1.073469210183482e-05, "loss": 2.6, "step": 6250 }, { "epoch": 0.7825, "grad_norm": 0.5859375, "learning_rate": 1.0617482854633593e-05, "loss": 2.5875, "step": 6260 }, { "epoch": 0.78375, "grad_norm": 0.6484375, "learning_rate": 1.0500831367760758e-05, "loss": 2.6, "step": 6270 }, { "epoch": 0.785, "grad_norm": 0.62109375, "learning_rate": 1.0384739533560947e-05, "loss": 2.6, "step": 6280 }, { "epoch": 0.78625, "grad_norm": 0.70703125, "learning_rate": 1.026920923530001e-05, "loss": 2.6125, "step": 6290 }, { "epoch": 0.7875, "grad_norm": 0.65234375, "learning_rate": 1.015424234713438e-05, "loss": 2.6, "step": 6300 }, { "epoch": 0.78875, "grad_norm": 0.7109375, "learning_rate": 1.0039840734080815e-05, "loss": 2.65, "step": 6310 }, { "epoch": 0.79, "grad_norm": 0.671875, "learning_rate": 9.926006251985994e-06, "loss": 2.65, "step": 6320 }, { "epoch": 0.79125, "grad_norm": 0.62890625, "learning_rate": 9.81274074749652e-06, "loss": 2.5875, "step": 6330 }, { "epoch": 0.7925, "grad_norm": 0.6171875, "learning_rate": 9.700046058028924e-06, "loss": 2.625, "step": 6340 }, { "epoch": 0.79375, "grad_norm": 0.6171875, "learning_rate": 9.587924011739825e-06, "loss": 2.525, "step": 6350 }, { "epoch": 0.795, "grad_norm": 0.67578125, "learning_rate": 9.476376427496343e-06, "loss": 2.625, "step": 6360 }, { "epoch": 0.79625, "grad_norm": 0.66015625, "learning_rate": 9.365405114846521e-06, "loss": 2.6, "step": 6370 }, { "epoch": 0.7975, "grad_norm": 0.65625, "learning_rate": 9.255011873990037e-06, "loss": 2.65, "step": 6380 }, { "epoch": 0.79875, "grad_norm": 0.65234375, "learning_rate": 9.145198495748956e-06, "loss": 2.6125, "step": 6390 }, { "epoch": 0.8, "grad_norm": 0.62109375, "learning_rate": 9.035966761538671e-06, "loss": 2.6375, "step": 6400 }, { "epoch": 0.80125, "grad_norm": 0.60546875, "learning_rate": 8.927318443339068e-06, "loss": 2.575, "step": 6410 }, { "epoch": 0.8025, "grad_norm": 0.62890625, "learning_rate": 8.819255303665685e-06, "loss": 2.5875, "step": 6420 }, { "epoch": 0.80375, "grad_norm": 0.640625, "learning_rate": 8.71177909554121e-06, "loss": 2.5875, "step": 6430 }, { "epoch": 0.805, "grad_norm": 0.6171875, "learning_rate": 8.604891562467016e-06, "loss": 2.55, "step": 6440 }, { "epoch": 0.80625, "grad_norm": 0.640625, "learning_rate": 8.498594438394815e-06, "loss": 2.65, "step": 6450 }, { "epoch": 0.8075, "grad_norm": 0.640625, "learning_rate": 8.392889447698626e-06, "loss": 2.65, "step": 6460 }, { "epoch": 0.80875, "grad_norm": 0.671875, "learning_rate": 8.287778305146745e-06, "loss": 2.625, "step": 6470 }, { "epoch": 0.81, "grad_norm": 0.66796875, "learning_rate": 8.183262715873937e-06, "loss": 2.6, "step": 6480 }, { "epoch": 0.81125, "grad_norm": 0.66015625, "learning_rate": 8.079344375353804e-06, "loss": 2.575, "step": 6490 }, { "epoch": 0.8125, "grad_norm": 0.6328125, "learning_rate": 7.976024969371195e-06, "loss": 2.5875, "step": 6500 }, { "epoch": 0.81375, "grad_norm": 0.640625, "learning_rate": 7.873306173994982e-06, "loss": 2.65, "step": 6510 }, { "epoch": 0.815, "grad_norm": 0.59765625, "learning_rate": 7.771189655550759e-06, "loss": 2.6125, "step": 6520 }, { "epoch": 0.81625, "grad_norm": 0.6484375, "learning_rate": 7.669677070593897e-06, "loss": 2.6625, "step": 6530 }, { "epoch": 0.8175, "grad_norm": 0.59765625, "learning_rate": 7.568770065882615e-06, "loss": 2.5125, "step": 6540 }, { "epoch": 0.81875, "grad_norm": 0.609375, "learning_rate": 7.468470278351275e-06, "loss": 2.65, "step": 6550 }, { "epoch": 0.82, "grad_norm": 0.625, "learning_rate": 7.368779335083855e-06, "loss": 2.5125, "step": 6560 }, { "epoch": 0.82125, "grad_norm": 0.6328125, "learning_rate": 7.269698853287509e-06, "loss": 2.625, "step": 6570 }, { "epoch": 0.8225, "grad_norm": 0.63671875, "learning_rate": 7.1712304402663905e-06, "loss": 2.625, "step": 6580 }, { "epoch": 0.82375, "grad_norm": 0.63671875, "learning_rate": 7.073375693395533e-06, "loss": 2.5375, "step": 6590 }, { "epoch": 0.825, "grad_norm": 0.65625, "learning_rate": 6.976136200094925e-06, "loss": 2.525, "step": 6600 }, { "epoch": 0.82625, "grad_norm": 0.65625, "learning_rate": 6.879513537803839e-06, "loss": 2.5, "step": 6610 }, { "epoch": 0.8275, "grad_norm": 0.60546875, "learning_rate": 6.78350927395512e-06, "loss": 2.625, "step": 6620 }, { "epoch": 0.82875, "grad_norm": 0.63671875, "learning_rate": 6.688124965949878e-06, "loss": 2.6, "step": 6630 }, { "epoch": 0.83, "grad_norm": 0.6171875, "learning_rate": 6.5933621611321606e-06, "loss": 2.625, "step": 6640 }, { "epoch": 0.83125, "grad_norm": 0.69921875, "learning_rate": 6.4992223967638356e-06, "loss": 2.575, "step": 6650 }, { "epoch": 0.8325, "grad_norm": 0.6171875, "learning_rate": 6.4057071999997e-06, "loss": 2.6625, "step": 6660 }, { "epoch": 0.83375, "grad_norm": 0.6015625, "learning_rate": 6.312818087862702e-06, "loss": 2.5625, "step": 6670 }, { "epoch": 0.835, "grad_norm": 0.6328125, "learning_rate": 6.220556567219269e-06, "loss": 2.55, "step": 6680 }, { "epoch": 0.83625, "grad_norm": 0.640625, "learning_rate": 6.128924134754954e-06, "loss": 2.5875, "step": 6690 }, { "epoch": 0.8375, "grad_norm": 0.625, "learning_rate": 6.037922276950077e-06, "loss": 2.625, "step": 6700 }, { "epoch": 0.83875, "grad_norm": 0.703125, "learning_rate": 5.947552470055675e-06, "loss": 2.6125, "step": 6710 }, { "epoch": 0.84, "grad_norm": 0.609375, "learning_rate": 5.8578161800694955e-06, "loss": 2.6, "step": 6720 }, { "epoch": 0.84125, "grad_norm": 0.65234375, "learning_rate": 5.768714862712274e-06, "loss": 2.625, "step": 6730 }, { "epoch": 0.8425, "grad_norm": 0.6484375, "learning_rate": 5.680249963404066e-06, "loss": 2.6375, "step": 6740 }, { "epoch": 0.84375, "grad_norm": 0.65625, "learning_rate": 5.592422917240817e-06, "loss": 2.5625, "step": 6750 }, { "epoch": 0.845, "grad_norm": 0.61328125, "learning_rate": 5.505235148971113e-06, "loss": 2.65, "step": 6760 }, { "epoch": 0.84625, "grad_norm": 0.6328125, "learning_rate": 5.4186880729730065e-06, "loss": 2.575, "step": 6770 }, { "epoch": 0.8475, "grad_norm": 0.640625, "learning_rate": 5.3327830932311294e-06, "loss": 2.6375, "step": 6780 }, { "epoch": 0.84875, "grad_norm": 0.5703125, "learning_rate": 5.247521603313908e-06, "loss": 2.6, "step": 6790 }, { "epoch": 0.85, "grad_norm": 0.61328125, "learning_rate": 5.162904986350903e-06, "loss": 2.6375, "step": 6800 }, { "epoch": 0.85125, "grad_norm": 0.59765625, "learning_rate": 5.078934615010453e-06, "loss": 2.6125, "step": 6810 }, { "epoch": 0.8525, "grad_norm": 0.54296875, "learning_rate": 4.9956118514773285e-06, "loss": 2.575, "step": 6820 }, { "epoch": 0.85375, "grad_norm": 0.61328125, "learning_rate": 4.912938047430702e-06, "loss": 2.6, "step": 6830 }, { "epoch": 0.855, "grad_norm": 0.66015625, "learning_rate": 4.830914544022191e-06, "loss": 2.5875, "step": 6840 }, { "epoch": 0.85625, "grad_norm": 0.6171875, "learning_rate": 4.749542671854065e-06, "loss": 2.6125, "step": 6850 }, { "epoch": 0.8575, "grad_norm": 0.59375, "learning_rate": 4.668823750957746e-06, "loss": 2.6125, "step": 6860 }, { "epoch": 0.85875, "grad_norm": 0.59375, "learning_rate": 4.588759090772301e-06, "loss": 2.6, "step": 6870 }, { "epoch": 0.86, "grad_norm": 0.5859375, "learning_rate": 4.509349990123271e-06, "loss": 2.6125, "step": 6880 }, { "epoch": 0.86125, "grad_norm": 0.6640625, "learning_rate": 4.43059773720157e-06, "loss": 2.55, "step": 6890 }, { "epoch": 0.8625, "grad_norm": 0.609375, "learning_rate": 4.352503609542582e-06, "loss": 2.6, "step": 6900 }, { "epoch": 0.86375, "grad_norm": 0.6484375, "learning_rate": 4.275068874005448e-06, "loss": 2.5875, "step": 6910 }, { "epoch": 0.865, "grad_norm": 0.6484375, "learning_rate": 4.1982947867525376e-06, "loss": 2.6, "step": 6920 }, { "epoch": 0.86625, "grad_norm": 0.6171875, "learning_rate": 4.122182593228998e-06, "loss": 2.6, "step": 6930 }, { "epoch": 0.8675, "grad_norm": 0.640625, "learning_rate": 4.046733528142659e-06, "loss": 2.575, "step": 6940 }, { "epoch": 0.86875, "grad_norm": 0.6171875, "learning_rate": 3.971948815443884e-06, "loss": 2.6, "step": 6950 }, { "epoch": 0.87, "grad_norm": 0.625, "learning_rate": 3.8978296683058134e-06, "loss": 2.575, "step": 6960 }, { "epoch": 0.87125, "grad_norm": 0.65625, "learning_rate": 3.824377289104635e-06, "loss": 2.6625, "step": 6970 }, { "epoch": 0.8725, "grad_norm": 0.640625, "learning_rate": 3.7515928694000868e-06, "loss": 2.6, "step": 6980 }, { "epoch": 0.87375, "grad_norm": 0.65234375, "learning_rate": 3.679477589916141e-06, "loss": 2.6, "step": 6990 }, { "epoch": 0.875, "grad_norm": 0.68359375, "learning_rate": 3.6080326205218047e-06, "loss": 2.625, "step": 7000 }, { "epoch": 0.87625, "grad_norm": 0.65234375, "learning_rate": 3.5372591202122173e-06, "loss": 2.5875, "step": 7010 }, { "epoch": 0.8775, "grad_norm": 0.65234375, "learning_rate": 3.4671582370897775e-06, "loss": 2.5625, "step": 7020 }, { "epoch": 0.87875, "grad_norm": 0.63671875, "learning_rate": 3.3977311083455567e-06, "loss": 2.575, "step": 7030 }, { "epoch": 0.88, "grad_norm": 0.671875, "learning_rate": 3.3289788602408633e-06, "loss": 2.575, "step": 7040 }, { "epoch": 0.88125, "grad_norm": 0.64453125, "learning_rate": 3.2609026080889098e-06, "loss": 2.625, "step": 7050 }, { "epoch": 0.8825, "grad_norm": 0.796875, "learning_rate": 3.193503456236805e-06, "loss": 2.5625, "step": 7060 }, { "epoch": 0.88375, "grad_norm": 0.640625, "learning_rate": 3.1267824980475686e-06, "loss": 2.6, "step": 7070 }, { "epoch": 0.885, "grad_norm": 0.62890625, "learning_rate": 3.0607408158824263e-06, "loss": 2.575, "step": 7080 }, { "epoch": 0.88625, "grad_norm": 0.640625, "learning_rate": 2.9953794810832686e-06, "loss": 2.5875, "step": 7090 }, { "epoch": 0.8875, "grad_norm": 0.58984375, "learning_rate": 2.930699553955209e-06, "loss": 2.5625, "step": 7100 }, { "epoch": 0.88875, "grad_norm": 0.64453125, "learning_rate": 2.8667020837494734e-06, "loss": 2.525, "step": 7110 }, { "epoch": 0.89, "grad_norm": 0.63671875, "learning_rate": 2.8033881086462838e-06, "loss": 2.55, "step": 7120 }, { "epoch": 0.89125, "grad_norm": 0.70703125, "learning_rate": 2.740758655738091e-06, "loss": 2.6125, "step": 7130 }, { "epoch": 0.8925, "grad_norm": 0.66015625, "learning_rate": 2.678814741012884e-06, "loss": 2.6375, "step": 7140 }, { "epoch": 0.89375, "grad_norm": 0.62109375, "learning_rate": 2.617557369337692e-06, "loss": 2.6, "step": 7150 }, { "epoch": 0.895, "grad_norm": 0.65234375, "learning_rate": 2.556987534442314e-06, "loss": 2.625, "step": 7160 }, { "epoch": 0.89625, "grad_norm": 0.67578125, "learning_rate": 2.4971062189031896e-06, "loss": 2.6375, "step": 7170 }, { "epoch": 0.8975, "grad_norm": 0.64453125, "learning_rate": 2.437914394127434e-06, "loss": 2.6625, "step": 7180 }, { "epoch": 0.89875, "grad_norm": 0.6328125, "learning_rate": 2.379413020337124e-06, "loss": 2.6125, "step": 7190 }, { "epoch": 0.9, "grad_norm": 0.6875, "learning_rate": 2.321603046553666e-06, "loss": 2.5375, "step": 7200 }, { "epoch": 0.90125, "grad_norm": 0.609375, "learning_rate": 2.2644854105824833e-06, "loss": 2.55, "step": 7210 }, { "epoch": 0.9025, "grad_norm": 0.6171875, "learning_rate": 2.208061038997693e-06, "loss": 2.65, "step": 7220 }, { "epoch": 0.90375, "grad_norm": 0.65625, "learning_rate": 2.152330847127169e-06, "loss": 2.5875, "step": 7230 }, { "epoch": 0.905, "grad_norm": 0.6796875, "learning_rate": 2.097295739037658e-06, "loss": 2.6125, "step": 7240 }, { "epoch": 0.90625, "grad_norm": 0.62890625, "learning_rate": 2.0429566075200905e-06, "loss": 2.6375, "step": 7250 }, { "epoch": 0.9075, "grad_norm": 0.640625, "learning_rate": 1.989314334075145e-06, "loss": 2.6, "step": 7260 }, { "epoch": 0.90875, "grad_norm": 0.61328125, "learning_rate": 1.9363697888989102e-06, "loss": 2.55, "step": 7270 }, { "epoch": 0.91, "grad_norm": 0.6171875, "learning_rate": 1.8841238308687815e-06, "loss": 2.625, "step": 7280 }, { "epoch": 0.91125, "grad_norm": 0.6171875, "learning_rate": 1.8325773075295377e-06, "loss": 2.625, "step": 7290 }, { "epoch": 0.9125, "grad_norm": 0.625, "learning_rate": 1.7817310550795746e-06, "loss": 2.5875, "step": 7300 }, { "epoch": 0.91375, "grad_norm": 0.66015625, "learning_rate": 1.7315858983573547e-06, "loss": 2.5875, "step": 7310 }, { "epoch": 0.915, "grad_norm": 0.66015625, "learning_rate": 1.6821426508280131e-06, "loss": 2.6, "step": 7320 }, { "epoch": 0.91625, "grad_norm": 0.6640625, "learning_rate": 1.6334021145701734e-06, "loss": 2.6125, "step": 7330 }, { "epoch": 0.9175, "grad_norm": 0.6328125, "learning_rate": 1.5853650802629329e-06, "loss": 2.6, "step": 7340 }, { "epoch": 0.91875, "grad_norm": 0.59765625, "learning_rate": 1.5380323271730223e-06, "loss": 2.6375, "step": 7350 }, { "epoch": 0.92, "grad_norm": 0.6796875, "learning_rate": 1.4914046231421975e-06, "loss": 2.6375, "step": 7360 }, { "epoch": 0.92125, "grad_norm": 0.6328125, "learning_rate": 1.445482724574737e-06, "loss": 2.6375, "step": 7370 }, { "epoch": 0.9225, "grad_norm": 0.640625, "learning_rate": 1.4002673764252146e-06, "loss": 2.625, "step": 7380 }, { "epoch": 0.92375, "grad_norm": 0.6484375, "learning_rate": 1.355759312186396e-06, "loss": 2.5875, "step": 7390 }, { "epoch": 0.925, "grad_norm": 0.6328125, "learning_rate": 1.3119592538773313e-06, "loss": 2.625, "step": 7400 }, { "epoch": 0.92625, "grad_norm": 0.7578125, "learning_rate": 1.268867912031652e-06, "loss": 2.65, "step": 7410 }, { "epoch": 0.9275, "grad_norm": 0.66796875, "learning_rate": 1.2264859856860623e-06, "loss": 2.6375, "step": 7420 }, { "epoch": 0.92875, "grad_norm": 0.625, "learning_rate": 1.1848141623689508e-06, "loss": 2.5875, "step": 7430 }, { "epoch": 0.93, "grad_norm": 0.64453125, "learning_rate": 1.1438531180892914e-06, "loss": 2.5875, "step": 7440 }, { "epoch": 0.93125, "grad_norm": 0.69921875, "learning_rate": 1.1036035173256366e-06, "loss": 2.625, "step": 7450 }, { "epoch": 0.9325, "grad_norm": 0.6484375, "learning_rate": 1.0640660130153802e-06, "loss": 2.6, "step": 7460 }, { "epoch": 0.93375, "grad_norm": 0.6484375, "learning_rate": 1.0252412465441018e-06, "loss": 2.6125, "step": 7470 }, { "epoch": 0.935, "grad_norm": 0.62109375, "learning_rate": 9.871298477352354e-07, "loss": 2.5625, "step": 7480 }, { "epoch": 0.93625, "grad_norm": 0.63671875, "learning_rate": 9.497324348397957e-07, "loss": 2.6125, "step": 7490 }, { "epoch": 0.9375, "grad_norm": 0.62890625, "learning_rate": 9.130496145263734e-07, "loss": 2.575, "step": 7500 }, { "epoch": 0.93875, "grad_norm": 0.6640625, "learning_rate": 8.770819818713016e-07, "loss": 2.6375, "step": 7510 }, { "epoch": 0.94, "grad_norm": 0.64453125, "learning_rate": 8.418301203489692e-07, "loss": 2.575, "step": 7520 }, { "epoch": 0.94125, "grad_norm": 0.62890625, "learning_rate": 8.072946018223932e-07, "loss": 2.55, "step": 7530 }, { "epoch": 0.9425, "grad_norm": 0.62109375, "learning_rate": 7.734759865339308e-07, "loss": 2.6125, "step": 7540 }, { "epoch": 0.94375, "grad_norm": 0.63671875, "learning_rate": 7.403748230961627e-07, "loss": 2.575, "step": 7550 }, { "epoch": 0.945, "grad_norm": 0.5859375, "learning_rate": 7.07991648483044e-07, "loss": 2.6125, "step": 7560 }, { "epoch": 0.94625, "grad_norm": 0.65234375, "learning_rate": 6.763269880211471e-07, "loss": 2.5875, "step": 7570 }, { "epoch": 0.9475, "grad_norm": 0.65625, "learning_rate": 6.453813553811677e-07, "loss": 2.5625, "step": 7580 }, { "epoch": 0.94875, "grad_norm": 0.6484375, "learning_rate": 6.151552525695825e-07, "loss": 2.5625, "step": 7590 }, { "epoch": 0.95, "grad_norm": 0.6171875, "learning_rate": 5.856491699204997e-07, "loss": 2.625, "step": 7600 }, { "epoch": 0.95125, "grad_norm": 0.60546875, "learning_rate": 5.568635860877158e-07, "loss": 2.55, "step": 7610 }, { "epoch": 0.9525, "grad_norm": 0.63671875, "learning_rate": 5.28798968036947e-07, "loss": 2.5125, "step": 7620 }, { "epoch": 0.95375, "grad_norm": 0.59765625, "learning_rate": 5.01455771038255e-07, "loss": 2.55, "step": 7630 }, { "epoch": 0.955, "grad_norm": 0.6171875, "learning_rate": 4.7483443865864787e-07, "loss": 2.65, "step": 7640 }, { "epoch": 0.95625, "grad_norm": 0.62109375, "learning_rate": 4.4893540275491085e-07, "loss": 2.575, "step": 7650 }, { "epoch": 0.9575, "grad_norm": 0.6640625, "learning_rate": 4.2375908346657723e-07, "loss": 2.5875, "step": 7660 }, { "epoch": 0.95875, "grad_norm": 0.625, "learning_rate": 3.993058892091234e-07, "loss": 2.6, "step": 7670 }, { "epoch": 0.96, "grad_norm": 0.6328125, "learning_rate": 3.7557621666733434e-07, "loss": 2.5875, "step": 7680 }, { "epoch": 0.96125, "grad_norm": 0.7109375, "learning_rate": 3.5257045078889385e-07, "loss": 2.5375, "step": 7690 }, { "epoch": 0.9625, "grad_norm": 0.6015625, "learning_rate": 3.3028896477810435e-07, "loss": 2.625, "step": 7700 }, { "epoch": 0.96375, "grad_norm": 0.640625, "learning_rate": 3.0873212008985685e-07, "loss": 2.575, "step": 7710 }, { "epoch": 0.965, "grad_norm": 0.66015625, "learning_rate": 2.8790026642376067e-07, "loss": 2.5875, "step": 7720 }, { "epoch": 0.96625, "grad_norm": 0.6171875, "learning_rate": 2.6779374171846283e-07, "loss": 2.6375, "step": 7730 }, { "epoch": 0.9675, "grad_norm": 0.62890625, "learning_rate": 2.4841287214619257e-07, "loss": 2.575, "step": 7740 }, { "epoch": 0.96875, "grad_norm": 0.64453125, "learning_rate": 2.2975797210742555e-07, "loss": 2.575, "step": 7750 }, { "epoch": 0.97, "grad_norm": 0.625, "learning_rate": 2.118293442258279e-07, "loss": 2.6375, "step": 7760 }, { "epoch": 0.97125, "grad_norm": 0.640625, "learning_rate": 1.9462727934332015e-07, "loss": 2.5875, "step": 7770 }, { "epoch": 0.9725, "grad_norm": 0.640625, "learning_rate": 1.7815205651537104e-07, "loss": 2.55, "step": 7780 }, { "epoch": 0.97375, "grad_norm": 0.64453125, "learning_rate": 1.6240394300645612e-07, "loss": 2.5875, "step": 7790 }, { "epoch": 0.975, "grad_norm": 0.6328125, "learning_rate": 1.473831942857512e-07, "loss": 2.6, "step": 7800 }, { "epoch": 0.97625, "grad_norm": 0.6328125, "learning_rate": 1.3309005402295572e-07, "loss": 2.5375, "step": 7810 }, { "epoch": 0.9775, "grad_norm": 0.6171875, "learning_rate": 1.1952475408435581e-07, "loss": 2.5375, "step": 7820 }, { "epoch": 0.97875, "grad_norm": 0.66015625, "learning_rate": 1.0668751452906245e-07, "loss": 2.625, "step": 7830 }, { "epoch": 0.98, "grad_norm": 0.61328125, "learning_rate": 9.457854360543917e-08, "loss": 2.5875, "step": 7840 }, { "epoch": 0.98125, "grad_norm": 0.71484375, "learning_rate": 8.319803774772484e-08, "loss": 2.65, "step": 7850 }, { "epoch": 0.9825, "grad_norm": 0.65234375, "learning_rate": 7.25461815728512e-08, "loss": 2.5625, "step": 7860 }, { "epoch": 0.98375, "grad_norm": 0.65234375, "learning_rate": 6.262314787743528e-08, "loss": 2.6, "step": 7870 }, { "epoch": 0.985, "grad_norm": 0.609375, "learning_rate": 5.34290976349916e-08, "loss": 2.625, "step": 7880 }, { "epoch": 0.98625, "grad_norm": 0.65234375, "learning_rate": 4.496417999330926e-08, "loss": 2.625, "step": 7890 }, { "epoch": 0.9875, "grad_norm": 0.6328125, "learning_rate": 3.7228532272043926e-08, "loss": 2.5875, "step": 7900 }, { "epoch": 0.98875, "grad_norm": 0.63671875, "learning_rate": 3.0222279960469556e-08, "loss": 2.625, "step": 7910 }, { "epoch": 0.99, "grad_norm": 0.66796875, "learning_rate": 2.3945536715470062e-08, "loss": 2.575, "step": 7920 }, { "epoch": 0.99125, "grad_norm": 0.6640625, "learning_rate": 1.8398404359680765e-08, "loss": 2.65, "step": 7930 }, { "epoch": 0.9925, "grad_norm": 0.6328125, "learning_rate": 1.3580972879824738e-08, "loss": 2.6125, "step": 7940 }, { "epoch": 0.99375, "grad_norm": 0.62890625, "learning_rate": 9.493320425283947e-09, "loss": 2.5625, "step": 7950 }, { "epoch": 0.995, "grad_norm": 0.6640625, "learning_rate": 6.135513306800289e-09, "loss": 2.5125, "step": 7960 }, { "epoch": 0.99625, "grad_norm": 0.64453125, "learning_rate": 3.507605995421437e-09, "loss": 2.5875, "step": 7970 }, { "epoch": 0.9975, "grad_norm": 0.671875, "learning_rate": 1.6096411215965612e-09, "loss": 2.6, "step": 7980 }, { "epoch": 0.99875, "grad_norm": 0.64453125, "learning_rate": 4.416494745168587e-10, "loss": 2.5875, "step": 7990 }, { "epoch": 1.0, "grad_norm": 0.625, "learning_rate": 3.6500015759832755e-12, "loss": 2.6, "step": 8000 } ], "logging_steps": 10, "max_steps": 8000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 100, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 5.35807370723328e+17, "train_batch_size": 16, "trial_name": null, "trial_params": null }