{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 22.0,
  "global_step": 79002,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 5e-09,
      "loss": 10.5192,
      "step": 1
    },
    {
      "epoch": 0.14,
      "learning_rate": 2.5e-06,
      "loss": 9.4826,
      "step": 500
    },
    {
      "epoch": 0.28,
      "learning_rate": 5e-06,
      "loss": 7.9029,
      "step": 1000
    },
    {
      "epoch": 0.42,
      "learning_rate": 7.5e-06,
      "loss": 6.8971,
      "step": 1500
    },
    {
      "epoch": 0.56,
      "learning_rate": 1e-05,
      "loss": 6.6133,
      "step": 2000
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.25e-05,
      "loss": 6.4578,
      "step": 2500
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.5e-05,
      "loss": 6.3502,
      "step": 3000
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.75e-05,
      "loss": 6.2558,
      "step": 3500
    },
    {
      "epoch": 1.11,
      "learning_rate": 2e-05,
      "loss": 6.1784,
      "step": 4000
    },
    {
      "epoch": 1.25,
      "learning_rate": 2.25e-05,
      "loss": 6.1171,
      "step": 4500
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.5e-05,
      "loss": 6.0613,
      "step": 5000
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 6.0146,
      "step": 5500
    },
    {
      "epoch": 1.67,
      "learning_rate": 3e-05,
      "loss": 5.9732,
      "step": 6000
    },
    {
      "epoch": 1.81,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 5.9385,
      "step": 6500
    },
    {
      "epoch": 1.95,
      "learning_rate": 3.5e-05,
      "loss": 5.9006,
      "step": 7000
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 5.8725,
      "step": 7500
    },
    {
      "epoch": 2.23,
      "learning_rate": 4e-05,
      "loss": 5.8389,
      "step": 8000
    },
    {
      "epoch": 2.37,
      "learning_rate": 4.2495e-05,
      "loss": 5.8149,
      "step": 8500
    },
    {
      "epoch": 2.51,
      "learning_rate": 4.4995000000000005e-05,
      "loss": 5.7907,
      "step": 9000
    },
    {
      "epoch": 2.65,
      "learning_rate": 4.7495e-05,
      "loss": 5.771,
      "step": 9500
    },
    {
      "epoch": 2.78,
      "learning_rate": 4.9995000000000005e-05,
      "loss": 5.7541,
      "step": 10000
    },
    {
      "epoch": 2.92,
      "learning_rate": 4.998254346606843e-05,
      "loss": 5.7367,
      "step": 10500
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.9965016825574876e-05,
      "loss": 5.7192,
      "step": 11000
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.994749018508132e-05,
      "loss": 5.7026,
      "step": 11500
    },
    {
      "epoch": 3.34,
      "learning_rate": 4.9929963544587775e-05,
      "loss": 5.6922,
      "step": 12000
    },
    {
      "epoch": 3.48,
      "learning_rate": 4.991247195737521e-05,
      "loss": 5.6796,
      "step": 12500
    },
    {
      "epoch": 3.62,
      "learning_rate": 4.9894945316881665e-05,
      "loss": 5.67,
      "step": 13000
    },
    {
      "epoch": 3.76,
      "learning_rate": 4.987741867638811e-05,
      "loss": 5.6599,
      "step": 13500
    },
    {
      "epoch": 3.9,
      "learning_rate": 4.9859892035894564e-05,
      "loss": 5.6487,
      "step": 14000
    },
    {
      "epoch": 4.04,
      "learning_rate": 4.9842400448681994e-05,
      "loss": 5.6437,
      "step": 14500
    },
    {
      "epoch": 4.18,
      "learning_rate": 4.982487380818845e-05,
      "loss": 5.6326,
      "step": 15000
    },
    {
      "epoch": 4.32,
      "learning_rate": 4.98073471676949e-05,
      "loss": 5.6252,
      "step": 15500
    },
    {
      "epoch": 4.46,
      "learning_rate": 4.978982052720135e-05,
      "loss": 5.6177,
      "step": 16000
    },
    {
      "epoch": 4.59,
      "learning_rate": 4.9772328939988783e-05,
      "loss": 5.6131,
      "step": 16500
    },
    {
      "epoch": 4.73,
      "learning_rate": 4.975483735277622e-05,
      "loss": 5.6043,
      "step": 17000
    },
    {
      "epoch": 4.87,
      "learning_rate": 4.9737310712282666e-05,
      "loss": 5.5996,
      "step": 17500
    },
    {
      "epoch": 5.01,
      "learning_rate": 4.971978407178912e-05,
      "loss": 5.5909,
      "step": 18000
    },
    {
      "epoch": 5.15,
      "learning_rate": 4.970225743129557e-05,
      "loss": 5.5874,
      "step": 18500
    },
    {
      "epoch": 5.29,
      "learning_rate": 4.968473079080202e-05,
      "loss": 5.5822,
      "step": 19000
    },
    {
      "epoch": 5.43,
      "learning_rate": 4.966723920358946e-05,
      "loss": 5.5773,
      "step": 19500
    },
    {
      "epoch": 5.57,
      "learning_rate": 4.964971256309591e-05,
      "loss": 5.5722,
      "step": 20000
    },
    {
      "epoch": 5.71,
      "learning_rate": 4.963218592260236e-05,
      "loss": 5.5707,
      "step": 20500
    },
    {
      "epoch": 5.85,
      "learning_rate": 4.961465928210881e-05,
      "loss": 5.5626,
      "step": 21000
    },
    {
      "epoch": 5.99,
      "learning_rate": 4.9597132641615254e-05,
      "loss": 5.5599,
      "step": 21500
    },
    {
      "epoch": 6.13,
      "learning_rate": 4.957964105440269e-05,
      "loss": 5.5536,
      "step": 22000
    },
    {
      "epoch": 6.27,
      "learning_rate": 4.9562114413909144e-05,
      "loss": 5.5496,
      "step": 22500
    },
    {
      "epoch": 6.4,
      "learning_rate": 4.95445877734156e-05,
      "loss": 5.5485,
      "step": 23000
    },
    {
      "epoch": 6.54,
      "learning_rate": 4.952706113292204e-05,
      "loss": 5.5436,
      "step": 23500
    },
    {
      "epoch": 6.68,
      "learning_rate": 4.950956954570948e-05,
      "loss": 5.5428,
      "step": 24000
    },
    {
      "epoch": 6.82,
      "learning_rate": 4.9492042905215926e-05,
      "loss": 5.5355,
      "step": 24500
    },
    {
      "epoch": 6.96,
      "learning_rate": 4.947451626472238e-05,
      "loss": 5.5348,
      "step": 25000
    },
    {
      "epoch": 7.1,
      "learning_rate": 4.9456989624228826e-05,
      "loss": 5.5275,
      "step": 25500
    },
    {
      "epoch": 7.24,
      "learning_rate": 4.943949803701627e-05,
      "loss": 5.5286,
      "step": 26000
    },
    {
      "epoch": 7.38,
      "learning_rate": 4.9421971396522716e-05,
      "loss": 5.5239,
      "step": 26500
    },
    {
      "epoch": 7.52,
      "learning_rate": 4.940444475602917e-05,
      "loss": 5.5223,
      "step": 27000
    },
    {
      "epoch": 7.66,
      "learning_rate": 4.9386953168816605e-05,
      "loss": 5.52,
      "step": 27500
    },
    {
      "epoch": 7.8,
      "learning_rate": 4.936942652832305e-05,
      "loss": 5.5177,
      "step": 28000
    },
    {
      "epoch": 7.94,
      "learning_rate": 4.9351899887829505e-05,
      "loss": 5.5147,
      "step": 28500
    },
    {
      "epoch": 8.08,
      "learning_rate": 4.933437324733595e-05,
      "loss": 5.5059,
      "step": 29000
    },
    {
      "epoch": 8.21,
      "learning_rate": 4.9316846606842404e-05,
      "loss": 5.5046,
      "step": 29500
    },
    {
      "epoch": 8.35,
      "learning_rate": 4.929931996634885e-05,
      "loss": 5.5033,
      "step": 30000
    },
    {
      "epoch": 8.49,
      "learning_rate": 4.92817933258553e-05,
      "loss": 5.5019,
      "step": 30500
    },
    {
      "epoch": 8.63,
      "learning_rate": 4.9264266685361756e-05,
      "loss": 5.4969,
      "step": 31000
    },
    {
      "epoch": 8.77,
      "learning_rate": 4.92467400448682e-05,
      "loss": 5.4955,
      "step": 31500
    },
    {
      "epoch": 8.91,
      "learning_rate": 4.922924845765564e-05,
      "loss": 5.4978,
      "step": 32000
    },
    {
      "epoch": 9.05,
      "learning_rate": 4.9211721817162086e-05,
      "loss": 5.4951,
      "step": 32500
    },
    {
      "epoch": 9.19,
      "learning_rate": 4.919419517666854e-05,
      "loss": 5.4846,
      "step": 33000
    },
    {
      "epoch": 9.33,
      "learning_rate": 4.9176668536174985e-05,
      "loss": 5.4871,
      "step": 33500
    },
    {
      "epoch": 9.47,
      "learning_rate": 4.915917694896243e-05,
      "loss": 5.4838,
      "step": 34000
    },
    {
      "epoch": 9.61,
      "learning_rate": 4.9141650308468875e-05,
      "loss": 5.4853,
      "step": 34500
    },
    {
      "epoch": 9.75,
      "learning_rate": 4.912412366797533e-05,
      "loss": 5.4807,
      "step": 35000
    },
    {
      "epoch": 9.89,
      "learning_rate": 4.9106597027481774e-05,
      "loss": 5.4789,
      "step": 35500
    },
    {
      "epoch": 10.03,
      "learning_rate": 4.908910544026921e-05,
      "loss": 5.4754,
      "step": 36000
    },
    {
      "epoch": 10.16,
      "learning_rate": 4.9071578799775664e-05,
      "loss": 5.4706,
      "step": 36500
    },
    {
      "epoch": 10.3,
      "learning_rate": 4.905405215928211e-05,
      "loss": 5.332,
      "step": 37000
    },
    {
      "epoch": 10.44,
      "learning_rate": 4.903652551878856e-05,
      "loss": 5.1202,
      "step": 37500
    },
    {
      "epoch": 10.58,
      "learning_rate": 4.901899887829501e-05,
      "loss": 4.9543,
      "step": 38000
    },
    {
      "epoch": 10.72,
      "learning_rate": 4.900147223780146e-05,
      "loss": 4.7848,
      "step": 38500
    },
    {
      "epoch": 10.86,
      "learning_rate": 4.898394559730791e-05,
      "loss": 4.6248,
      "step": 39000
    },
    {
      "epoch": 11.0,
      "learning_rate": 4.8966418956814355e-05,
      "loss": 4.4711,
      "step": 39500
    },
    {
      "epoch": 11.14,
      "learning_rate": 4.894889231632081e-05,
      "loss": 4.3324,
      "step": 40000
    },
    {
      "epoch": 11.28,
      "learning_rate": 4.8931400729108245e-05,
      "loss": 4.1831,
      "step": 40500
    },
    {
      "epoch": 11.42,
      "learning_rate": 4.89138740886147e-05,
      "loss": 3.845,
      "step": 41000
    },
    {
      "epoch": 11.56,
      "learning_rate": 4.889634744812115e-05,
      "loss": 3.3852,
      "step": 41500
    },
    {
      "epoch": 11.7,
      "learning_rate": 4.88788208076276e-05,
      "loss": 2.749,
      "step": 42000
    },
    {
      "epoch": 11.84,
      "learning_rate": 4.886136427369602e-05,
      "loss": 2.4339,
      "step": 42500
    },
    {
      "epoch": 11.97,
      "learning_rate": 4.884383763320247e-05,
      "loss": 2.2505,
      "step": 43000
    },
    {
      "epoch": 12.11,
      "learning_rate": 4.882631099270892e-05,
      "loss": 2.1255,
      "step": 43500
    },
    {
      "epoch": 12.25,
      "learning_rate": 4.880878435221537e-05,
      "loss": 2.037,
      "step": 44000
    },
    {
      "epoch": 12.39,
      "learning_rate": 4.879125771172182e-05,
      "loss": 1.9662,
      "step": 44500
    },
    {
      "epoch": 12.53,
      "learning_rate": 4.877376612450926e-05,
      "loss": 1.909,
      "step": 45000
    },
    {
      "epoch": 12.67,
      "learning_rate": 4.8756239484015706e-05,
      "loss": 1.8613,
      "step": 45500
    },
    {
      "epoch": 12.81,
      "learning_rate": 4.873871284352215e-05,
      "loss": 1.8191,
      "step": 46000
    },
    {
      "epoch": 12.95,
      "learning_rate": 4.8721186203028606e-05,
      "loss": 1.7808,
      "step": 46500
    },
    {
      "epoch": 13.09,
      "learning_rate": 4.870369461581604e-05,
      "loss": 1.745,
      "step": 47000
    },
    {
      "epoch": 13.23,
      "learning_rate": 4.8686167975322495e-05,
      "loss": 1.7123,
      "step": 47500
    },
    {
      "epoch": 13.37,
      "learning_rate": 4.866864133482894e-05,
      "loss": 1.6834,
      "step": 48000
    },
    {
      "epoch": 13.51,
      "learning_rate": 4.8651114694335395e-05,
      "loss": 1.6541,
      "step": 48500
    },
    {
      "epoch": 13.65,
      "learning_rate": 4.8633623107122825e-05,
      "loss": 1.6308,
      "step": 49000
    },
    {
      "epoch": 13.78,
      "learning_rate": 4.861609646662928e-05,
      "loss": 1.6093,
      "step": 49500
    },
    {
      "epoch": 13.92,
      "learning_rate": 4.859856982613573e-05,
      "loss": 1.5854,
      "step": 50000
    },
    {
      "epoch": 14.06,
      "learning_rate": 4.858104318564218e-05,
      "loss": 1.5645,
      "step": 50500
    },
    {
      "epoch": 14.2,
      "learning_rate": 4.856351654514863e-05,
      "loss": 1.5449,
      "step": 51000
    },
    {
      "epoch": 14.34,
      "learning_rate": 4.854602495793607e-05,
      "loss": 1.5306,
      "step": 51500
    },
    {
      "epoch": 14.48,
      "learning_rate": 4.852849831744252e-05,
      "loss": 1.5134,
      "step": 52000
    },
    {
      "epoch": 14.62,
      "learning_rate": 4.8510971676948966e-05,
      "loss": 1.4979,
      "step": 52500
    },
    {
      "epoch": 14.76,
      "learning_rate": 4.849344503645541e-05,
      "loss": 1.4835,
      "step": 53000
    },
    {
      "epoch": 14.9,
      "learning_rate": 4.847595344924285e-05,
      "loss": 1.468,
      "step": 53500
    },
    {
      "epoch": 15.04,
      "learning_rate": 4.84584268087493e-05,
      "loss": 1.4542,
      "step": 54000
    },
    {
      "epoch": 15.18,
      "learning_rate": 4.844090016825575e-05,
      "loss": 1.4416,
      "step": 54500
    },
    {
      "epoch": 15.32,
      "learning_rate": 4.842340858104319e-05,
      "loss": 1.4277,
      "step": 55000
    },
    {
      "epoch": 15.46,
      "learning_rate": 4.840588194054964e-05,
      "loss": 1.4185,
      "step": 55500
    },
    {
      "epoch": 15.59,
      "learning_rate": 4.8388355300056085e-05,
      "loss": 1.405,
      "step": 56000
    },
    {
      "epoch": 15.73,
      "learning_rate": 4.837082865956254e-05,
      "loss": 1.3956,
      "step": 56500
    },
    {
      "epoch": 15.87,
      "learning_rate": 4.8353302019068984e-05,
      "loss": 1.384,
      "step": 57000
    },
    {
      "epoch": 16.01,
      "learning_rate": 4.833581043185643e-05,
      "loss": 1.3745,
      "step": 57500
    },
    {
      "epoch": 16.15,
      "learning_rate": 4.8318283791362874e-05,
      "loss": 1.3639,
      "step": 58000
    },
    {
      "epoch": 16.29,
      "learning_rate": 4.830075715086933e-05,
      "loss": 1.3543,
      "step": 58500
    },
    {
      "epoch": 16.43,
      "learning_rate": 4.828323051037577e-05,
      "loss": 1.347,
      "step": 59000
    },
    {
      "epoch": 16.57,
      "learning_rate": 4.8265703869882226e-05,
      "loss": 1.3377,
      "step": 59500
    },
    {
      "epoch": 16.71,
      "learning_rate": 4.8248212282669656e-05,
      "loss": 1.3291,
      "step": 60000
    },
    {
      "epoch": 16.85,
      "learning_rate": 4.823068564217611e-05,
      "loss": 1.3214,
      "step": 60500
    },
    {
      "epoch": 16.99,
      "learning_rate": 4.821315900168256e-05,
      "loss": 1.3146,
      "step": 61000
    },
    {
      "epoch": 17.13,
      "learning_rate": 4.819563236118901e-05,
      "loss": 1.3041,
      "step": 61500
    },
    {
      "epoch": 17.27,
      "learning_rate": 4.817810572069546e-05,
      "loss": 1.2975,
      "step": 62000
    },
    {
      "epoch": 17.4,
      "learning_rate": 4.8160579080201915e-05,
      "loss": 1.2908,
      "step": 62500
    },
    {
      "epoch": 17.54,
      "learning_rate": 4.814305243970836e-05,
      "loss": 1.2835,
      "step": 63000
    },
    {
      "epoch": 17.68,
      "learning_rate": 4.812556085249579e-05,
      "loss": 1.2766,
      "step": 63500
    },
    {
      "epoch": 17.82,
      "learning_rate": 4.8108034212002244e-05,
      "loss": 1.2715,
      "step": 64000
    },
    {
      "epoch": 17.96,
      "learning_rate": 4.80905075715087e-05,
      "loss": 1.2655,
      "step": 64500
    },
    {
      "epoch": 18.1,
      "learning_rate": 4.807298093101514e-05,
      "loss": 1.2565,
      "step": 65000
    },
    {
      "epoch": 18.24,
      "learning_rate": 4.805548934380259e-05,
      "loss": 1.2506,
      "step": 65500
    },
    {
      "epoch": 18.38,
      "learning_rate": 4.803796270330903e-05,
      "loss": 1.2443,
      "step": 66000
    },
    {
      "epoch": 18.52,
      "learning_rate": 4.802043606281548e-05,
      "loss": 1.2386,
      "step": 66500
    },
    {
      "epoch": 18.66,
      "learning_rate": 4.800290942232193e-05,
      "loss": 1.2338,
      "step": 67000
    },
    {
      "epoch": 18.8,
      "learning_rate": 4.798541783510937e-05,
      "loss": 1.2287,
      "step": 67500
    },
    {
      "epoch": 18.94,
      "learning_rate": 4.7967926247896806e-05,
      "loss": 1.2233,
      "step": 68000
    },
    {
      "epoch": 19.08,
      "learning_rate": 4.795039960740326e-05,
      "loss": 1.2185,
      "step": 68500
    },
    {
      "epoch": 19.21,
      "learning_rate": 4.7932872966909705e-05,
      "loss": 1.2133,
      "step": 69000
    },
    {
      "epoch": 19.35,
      "learning_rate": 4.791534632641616e-05,
      "loss": 1.2088,
      "step": 69500
    },
    {
      "epoch": 19.49,
      "learning_rate": 4.7897819685922604e-05,
      "loss": 1.2036,
      "step": 70000
    },
    {
      "epoch": 19.63,
      "learning_rate": 4.788029304542905e-05,
      "loss": 1.1979,
      "step": 70500
    },
    {
      "epoch": 19.77,
      "learning_rate": 4.7862766404935504e-05,
      "loss": 1.1919,
      "step": 71000
    },
    {
      "epoch": 19.91,
      "learning_rate": 4.784527481772294e-05,
      "loss": 1.1893,
      "step": 71500
    },
    {
      "epoch": 20.05,
      "learning_rate": 4.7827748177229394e-05,
      "loss": 1.1863,
      "step": 72000
    },
    {
      "epoch": 20.19,
      "learning_rate": 4.781022153673584e-05,
      "loss": 1.1801,
      "step": 72500
    },
    {
      "epoch": 20.33,
      "learning_rate": 4.779269489624229e-05,
      "loss": 1.1777,
      "step": 73000
    },
    {
      "epoch": 20.47,
      "learning_rate": 4.777516825574874e-05,
      "loss": 1.1714,
      "step": 73500
    },
    {
      "epoch": 20.61,
      "learning_rate": 4.7757676668536176e-05,
      "loss": 1.1684,
      "step": 74000
    },
    {
      "epoch": 20.75,
      "learning_rate": 4.774015002804263e-05,
      "loss": 1.165,
      "step": 74500
    },
    {
      "epoch": 20.89,
      "learning_rate": 4.7722623387549075e-05,
      "loss": 1.1602,
      "step": 75000
    },
    {
      "epoch": 21.02,
      "learning_rate": 4.770509674705553e-05,
      "loss": 1.1567,
      "step": 75500
    },
    {
      "epoch": 21.16,
      "learning_rate": 4.7687570106561975e-05,
      "loss": 1.1524,
      "step": 76000
    },
    {
      "epoch": 21.3,
      "learning_rate": 4.767007851934942e-05,
      "loss": 1.1502,
      "step": 76500
    },
    {
      "epoch": 21.44,
      "learning_rate": 4.7652551878855864e-05,
      "loss": 1.1442,
      "step": 77000
    },
    {
      "epoch": 21.58,
      "learning_rate": 4.763502523836231e-05,
      "loss": 1.1422,
      "step": 77500
    },
    {
      "epoch": 21.72,
      "learning_rate": 4.7617498597868764e-05,
      "loss": 1.1384,
      "step": 78000
    },
    {
      "epoch": 21.86,
      "learning_rate": 4.759997195737521e-05,
      "loss": 1.1361,
      "step": 78500
    },
    {
      "epoch": 22.0,
      "learning_rate": 4.758244531688166e-05,
      "loss": 1.1318,
      "step": 79000
    }
  ],
  "max_steps": 1436400,
  "num_train_epochs": 400,
  "total_flos": 2.129238171230562e+19,
  "trial_name": null,
  "trial_params": null
}