Harish Tayyar Madabushi
added model
c90196a
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"global_step": 250000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 1.9980000000000002e-05,
"loss": 1.5856,
"step": 500
},
{
"epoch": 0.02,
"learning_rate": 1.9960000000000002e-05,
"loss": 1.1851,
"step": 1000
},
{
"epoch": 0.03,
"learning_rate": 1.9940000000000002e-05,
"loss": 1.1964,
"step": 1500
},
{
"epoch": 0.04,
"learning_rate": 1.9920000000000002e-05,
"loss": 1.1492,
"step": 2000
},
{
"epoch": 0.05,
"learning_rate": 1.9900000000000003e-05,
"loss": 1.1113,
"step": 2500
},
{
"epoch": 0.06,
"learning_rate": 1.9880000000000003e-05,
"loss": 1.1244,
"step": 3000
},
{
"epoch": 0.07,
"learning_rate": 1.9860000000000003e-05,
"loss": 1.1124,
"step": 3500
},
{
"epoch": 0.08,
"learning_rate": 1.9840000000000003e-05,
"loss": 1.0842,
"step": 4000
},
{
"epoch": 0.09,
"learning_rate": 1.982e-05,
"loss": 1.0997,
"step": 4500
},
{
"epoch": 0.1,
"learning_rate": 1.98e-05,
"loss": 1.0618,
"step": 5000
},
{
"epoch": 0.11,
"learning_rate": 1.978e-05,
"loss": 1.1082,
"step": 5500
},
{
"epoch": 0.12,
"learning_rate": 1.976e-05,
"loss": 1.0901,
"step": 6000
},
{
"epoch": 0.13,
"learning_rate": 1.974e-05,
"loss": 1.1198,
"step": 6500
},
{
"epoch": 0.14,
"learning_rate": 1.972e-05,
"loss": 1.1091,
"step": 7000
},
{
"epoch": 0.15,
"learning_rate": 1.97e-05,
"loss": 1.0496,
"step": 7500
},
{
"epoch": 0.16,
"learning_rate": 1.968e-05,
"loss": 1.0579,
"step": 8000
},
{
"epoch": 0.17,
"learning_rate": 1.966e-05,
"loss": 1.0638,
"step": 8500
},
{
"epoch": 0.18,
"learning_rate": 1.9640000000000002e-05,
"loss": 1.0912,
"step": 9000
},
{
"epoch": 0.19,
"learning_rate": 1.9620000000000002e-05,
"loss": 1.0564,
"step": 9500
},
{
"epoch": 0.2,
"learning_rate": 1.9600000000000002e-05,
"loss": 1.0759,
"step": 10000
},
{
"epoch": 0.21,
"learning_rate": 1.9580000000000002e-05,
"loss": 1.0336,
"step": 10500
},
{
"epoch": 0.22,
"learning_rate": 1.9560000000000002e-05,
"loss": 1.0601,
"step": 11000
},
{
"epoch": 0.23,
"learning_rate": 1.9540000000000003e-05,
"loss": 1.0388,
"step": 11500
},
{
"epoch": 0.24,
"learning_rate": 1.9520000000000003e-05,
"loss": 1.0621,
"step": 12000
},
{
"epoch": 0.25,
"learning_rate": 1.95e-05,
"loss": 1.0859,
"step": 12500
},
{
"epoch": 0.26,
"learning_rate": 1.948e-05,
"loss": 1.0492,
"step": 13000
},
{
"epoch": 0.27,
"learning_rate": 1.946e-05,
"loss": 1.0551,
"step": 13500
},
{
"epoch": 0.28,
"learning_rate": 1.944e-05,
"loss": 1.0248,
"step": 14000
},
{
"epoch": 0.29,
"learning_rate": 1.942e-05,
"loss": 1.0416,
"step": 14500
},
{
"epoch": 0.3,
"learning_rate": 1.94e-05,
"loss": 1.0339,
"step": 15000
},
{
"epoch": 0.31,
"learning_rate": 1.938e-05,
"loss": 1.0452,
"step": 15500
},
{
"epoch": 0.32,
"learning_rate": 1.936e-05,
"loss": 1.0178,
"step": 16000
},
{
"epoch": 0.33,
"learning_rate": 1.934e-05,
"loss": 1.0517,
"step": 16500
},
{
"epoch": 0.34,
"learning_rate": 1.932e-05,
"loss": 1.0101,
"step": 17000
},
{
"epoch": 0.35,
"learning_rate": 1.93e-05,
"loss": 1.0358,
"step": 17500
},
{
"epoch": 0.36,
"learning_rate": 1.9280000000000002e-05,
"loss": 1.0145,
"step": 18000
},
{
"epoch": 0.37,
"learning_rate": 1.9260000000000002e-05,
"loss": 1.0449,
"step": 18500
},
{
"epoch": 0.38,
"learning_rate": 1.9240000000000002e-05,
"loss": 1.0334,
"step": 19000
},
{
"epoch": 0.39,
"learning_rate": 1.9220000000000002e-05,
"loss": 0.988,
"step": 19500
},
{
"epoch": 0.4,
"learning_rate": 1.9200000000000003e-05,
"loss": 1.0197,
"step": 20000
},
{
"epoch": 0.41,
"learning_rate": 1.918e-05,
"loss": 1.0102,
"step": 20500
},
{
"epoch": 0.42,
"learning_rate": 1.916e-05,
"loss": 1.0621,
"step": 21000
},
{
"epoch": 0.43,
"learning_rate": 1.914e-05,
"loss": 1.0236,
"step": 21500
},
{
"epoch": 0.44,
"learning_rate": 1.912e-05,
"loss": 1.0556,
"step": 22000
},
{
"epoch": 0.45,
"learning_rate": 1.91e-05,
"loss": 1.0212,
"step": 22500
},
{
"epoch": 0.46,
"learning_rate": 1.908e-05,
"loss": 0.9991,
"step": 23000
},
{
"epoch": 0.47,
"learning_rate": 1.906e-05,
"loss": 1.0431,
"step": 23500
},
{
"epoch": 0.48,
"learning_rate": 1.904e-05,
"loss": 1.0099,
"step": 24000
},
{
"epoch": 0.49,
"learning_rate": 1.902e-05,
"loss": 1.0235,
"step": 24500
},
{
"epoch": 0.5,
"learning_rate": 1.9e-05,
"loss": 1.0044,
"step": 25000
},
{
"epoch": 0.51,
"learning_rate": 1.898e-05,
"loss": 1.0476,
"step": 25500
},
{
"epoch": 0.52,
"learning_rate": 1.896e-05,
"loss": 1.0168,
"step": 26000
},
{
"epoch": 0.53,
"learning_rate": 1.894e-05,
"loss": 1.03,
"step": 26500
},
{
"epoch": 0.54,
"learning_rate": 1.8920000000000002e-05,
"loss": 1.035,
"step": 27000
},
{
"epoch": 0.55,
"learning_rate": 1.8900000000000002e-05,
"loss": 1.0196,
"step": 27500
},
{
"epoch": 0.56,
"learning_rate": 1.8880000000000002e-05,
"loss": 1.0446,
"step": 28000
},
{
"epoch": 0.57,
"learning_rate": 1.886e-05,
"loss": 1.0376,
"step": 28500
},
{
"epoch": 0.58,
"learning_rate": 1.884e-05,
"loss": 0.9865,
"step": 29000
},
{
"epoch": 0.59,
"learning_rate": 1.882e-05,
"loss": 1.0091,
"step": 29500
},
{
"epoch": 0.6,
"learning_rate": 1.88e-05,
"loss": 1.0268,
"step": 30000
},
{
"epoch": 0.61,
"learning_rate": 1.878e-05,
"loss": 0.999,
"step": 30500
},
{
"epoch": 0.62,
"learning_rate": 1.876e-05,
"loss": 0.9908,
"step": 31000
},
{
"epoch": 0.63,
"learning_rate": 1.8740000000000004e-05,
"loss": 1.0225,
"step": 31500
},
{
"epoch": 0.64,
"learning_rate": 1.8720000000000004e-05,
"loss": 1.0211,
"step": 32000
},
{
"epoch": 0.65,
"learning_rate": 1.8700000000000004e-05,
"loss": 1.0369,
"step": 32500
},
{
"epoch": 0.66,
"learning_rate": 1.8680000000000004e-05,
"loss": 0.9936,
"step": 33000
},
{
"epoch": 0.67,
"learning_rate": 1.866e-05,
"loss": 0.9871,
"step": 33500
},
{
"epoch": 0.68,
"learning_rate": 1.864e-05,
"loss": 1.0057,
"step": 34000
},
{
"epoch": 0.69,
"learning_rate": 1.862e-05,
"loss": 0.9983,
"step": 34500
},
{
"epoch": 0.7,
"learning_rate": 1.86e-05,
"loss": 1.0156,
"step": 35000
},
{
"epoch": 0.71,
"learning_rate": 1.858e-05,
"loss": 1.0123,
"step": 35500
},
{
"epoch": 0.72,
"learning_rate": 1.8560000000000002e-05,
"loss": 1.0401,
"step": 36000
},
{
"epoch": 0.73,
"learning_rate": 1.8540000000000002e-05,
"loss": 0.9898,
"step": 36500
},
{
"epoch": 0.74,
"learning_rate": 1.8520000000000002e-05,
"loss": 1.0112,
"step": 37000
},
{
"epoch": 0.75,
"learning_rate": 1.8500000000000002e-05,
"loss": 1.0024,
"step": 37500
},
{
"epoch": 0.76,
"learning_rate": 1.8480000000000003e-05,
"loss": 0.9794,
"step": 38000
},
{
"epoch": 0.77,
"learning_rate": 1.8460000000000003e-05,
"loss": 0.9986,
"step": 38500
},
{
"epoch": 0.78,
"learning_rate": 1.8440000000000003e-05,
"loss": 1.0181,
"step": 39000
},
{
"epoch": 0.79,
"learning_rate": 1.8420000000000003e-05,
"loss": 0.9949,
"step": 39500
},
{
"epoch": 0.8,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.9803,
"step": 40000
},
{
"epoch": 0.81,
"learning_rate": 1.8380000000000004e-05,
"loss": 1.0184,
"step": 40500
},
{
"epoch": 0.82,
"learning_rate": 1.8360000000000004e-05,
"loss": 0.9792,
"step": 41000
},
{
"epoch": 0.83,
"learning_rate": 1.834e-05,
"loss": 0.9938,
"step": 41500
},
{
"epoch": 0.84,
"learning_rate": 1.832e-05,
"loss": 1.0015,
"step": 42000
},
{
"epoch": 0.85,
"learning_rate": 1.83e-05,
"loss": 0.9809,
"step": 42500
},
{
"epoch": 0.86,
"learning_rate": 1.828e-05,
"loss": 0.9914,
"step": 43000
},
{
"epoch": 0.87,
"learning_rate": 1.826e-05,
"loss": 0.9457,
"step": 43500
},
{
"epoch": 0.88,
"learning_rate": 1.824e-05,
"loss": 1.0056,
"step": 44000
},
{
"epoch": 0.89,
"learning_rate": 1.8220000000000002e-05,
"loss": 1.0131,
"step": 44500
},
{
"epoch": 0.9,
"learning_rate": 1.8200000000000002e-05,
"loss": 1.0017,
"step": 45000
},
{
"epoch": 0.91,
"learning_rate": 1.8180000000000002e-05,
"loss": 0.9875,
"step": 45500
},
{
"epoch": 0.92,
"learning_rate": 1.8160000000000002e-05,
"loss": 1.004,
"step": 46000
},
{
"epoch": 0.93,
"learning_rate": 1.8140000000000003e-05,
"loss": 1.0072,
"step": 46500
},
{
"epoch": 0.94,
"learning_rate": 1.8120000000000003e-05,
"loss": 1.003,
"step": 47000
},
{
"epoch": 0.95,
"learning_rate": 1.8100000000000003e-05,
"loss": 0.9947,
"step": 47500
},
{
"epoch": 0.96,
"learning_rate": 1.8080000000000003e-05,
"loss": 0.9802,
"step": 48000
},
{
"epoch": 0.97,
"learning_rate": 1.8060000000000003e-05,
"loss": 0.9708,
"step": 48500
},
{
"epoch": 0.98,
"learning_rate": 1.8040000000000003e-05,
"loss": 1.0065,
"step": 49000
},
{
"epoch": 0.99,
"learning_rate": 1.802e-05,
"loss": 1.0028,
"step": 49500
},
{
"epoch": 1.0,
"learning_rate": 1.8e-05,
"loss": 0.9957,
"step": 50000
},
{
"epoch": 1.0,
"eval_loss": 1.0293926000595093,
"eval_runtime": 36.7211,
"eval_samples_per_second": 229.813,
"step": 50000
},
{
"epoch": 1.01,
"learning_rate": 1.798e-05,
"loss": 1.0013,
"step": 50500
},
{
"epoch": 1.02,
"learning_rate": 1.796e-05,
"loss": 0.9563,
"step": 51000
},
{
"epoch": 1.03,
"learning_rate": 1.794e-05,
"loss": 0.9488,
"step": 51500
},
{
"epoch": 1.04,
"learning_rate": 1.792e-05,
"loss": 0.9799,
"step": 52000
},
{
"epoch": 1.05,
"learning_rate": 1.79e-05,
"loss": 0.975,
"step": 52500
},
{
"epoch": 1.06,
"learning_rate": 1.788e-05,
"loss": 0.9706,
"step": 53000
},
{
"epoch": 1.07,
"learning_rate": 1.7860000000000002e-05,
"loss": 0.9602,
"step": 53500
},
{
"epoch": 1.08,
"learning_rate": 1.7840000000000002e-05,
"loss": 0.9734,
"step": 54000
},
{
"epoch": 1.09,
"learning_rate": 1.7820000000000002e-05,
"loss": 0.9597,
"step": 54500
},
{
"epoch": 1.1,
"learning_rate": 1.7800000000000002e-05,
"loss": 0.9637,
"step": 55000
},
{
"epoch": 1.11,
"learning_rate": 1.7780000000000003e-05,
"loss": 0.9735,
"step": 55500
},
{
"epoch": 1.12,
"learning_rate": 1.7760000000000003e-05,
"loss": 0.9829,
"step": 56000
},
{
"epoch": 1.13,
"learning_rate": 1.7740000000000003e-05,
"loss": 0.9762,
"step": 56500
},
{
"epoch": 1.14,
"learning_rate": 1.7720000000000003e-05,
"loss": 0.9713,
"step": 57000
},
{
"epoch": 1.15,
"learning_rate": 1.77e-05,
"loss": 0.9629,
"step": 57500
},
{
"epoch": 1.16,
"learning_rate": 1.768e-05,
"loss": 0.9717,
"step": 58000
},
{
"epoch": 1.17,
"learning_rate": 1.766e-05,
"loss": 0.9928,
"step": 58500
},
{
"epoch": 1.18,
"learning_rate": 1.764e-05,
"loss": 0.925,
"step": 59000
},
{
"epoch": 1.19,
"learning_rate": 1.762e-05,
"loss": 0.9672,
"step": 59500
},
{
"epoch": 1.2,
"learning_rate": 1.76e-05,
"loss": 0.9915,
"step": 60000
},
{
"epoch": 1.21,
"learning_rate": 1.758e-05,
"loss": 0.9888,
"step": 60500
},
{
"epoch": 1.22,
"learning_rate": 1.756e-05,
"loss": 0.9473,
"step": 61000
},
{
"epoch": 1.23,
"learning_rate": 1.754e-05,
"loss": 0.9552,
"step": 61500
},
{
"epoch": 1.24,
"learning_rate": 1.752e-05,
"loss": 0.9878,
"step": 62000
},
{
"epoch": 1.25,
"learning_rate": 1.7500000000000002e-05,
"loss": 0.972,
"step": 62500
},
{
"epoch": 1.26,
"learning_rate": 1.7480000000000002e-05,
"loss": 1.0064,
"step": 63000
},
{
"epoch": 1.27,
"learning_rate": 1.7460000000000002e-05,
"loss": 0.9613,
"step": 63500
},
{
"epoch": 1.28,
"learning_rate": 1.7440000000000002e-05,
"loss": 0.9594,
"step": 64000
},
{
"epoch": 1.29,
"learning_rate": 1.7420000000000003e-05,
"loss": 0.978,
"step": 64500
},
{
"epoch": 1.3,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.9826,
"step": 65000
},
{
"epoch": 1.31,
"learning_rate": 1.7380000000000003e-05,
"loss": 0.958,
"step": 65500
},
{
"epoch": 1.32,
"learning_rate": 1.736e-05,
"loss": 0.9604,
"step": 66000
},
{
"epoch": 1.33,
"learning_rate": 1.734e-05,
"loss": 0.9622,
"step": 66500
},
{
"epoch": 1.34,
"learning_rate": 1.732e-05,
"loss": 0.938,
"step": 67000
},
{
"epoch": 1.35,
"learning_rate": 1.73e-05,
"loss": 0.9648,
"step": 67500
},
{
"epoch": 1.36,
"learning_rate": 1.728e-05,
"loss": 0.9453,
"step": 68000
},
{
"epoch": 1.37,
"learning_rate": 1.726e-05,
"loss": 0.9513,
"step": 68500
},
{
"epoch": 1.38,
"learning_rate": 1.724e-05,
"loss": 0.9811,
"step": 69000
},
{
"epoch": 1.39,
"learning_rate": 1.722e-05,
"loss": 0.9797,
"step": 69500
},
{
"epoch": 1.4,
"learning_rate": 1.72e-05,
"loss": 0.9241,
"step": 70000
},
{
"epoch": 1.41,
"learning_rate": 1.718e-05,
"loss": 0.948,
"step": 70500
},
{
"epoch": 1.42,
"learning_rate": 1.7160000000000002e-05,
"loss": 0.9596,
"step": 71000
},
{
"epoch": 1.43,
"learning_rate": 1.7140000000000002e-05,
"loss": 0.9635,
"step": 71500
},
{
"epoch": 1.44,
"learning_rate": 1.7120000000000002e-05,
"loss": 0.983,
"step": 72000
},
{
"epoch": 1.45,
"learning_rate": 1.7100000000000002e-05,
"loss": 0.9788,
"step": 72500
},
{
"epoch": 1.46,
"learning_rate": 1.7080000000000002e-05,
"loss": 0.9827,
"step": 73000
},
{
"epoch": 1.47,
"learning_rate": 1.7060000000000003e-05,
"loss": 0.9305,
"step": 73500
},
{
"epoch": 1.48,
"learning_rate": 1.704e-05,
"loss": 0.9367,
"step": 74000
},
{
"epoch": 1.49,
"learning_rate": 1.702e-05,
"loss": 0.9379,
"step": 74500
},
{
"epoch": 1.5,
"learning_rate": 1.7e-05,
"loss": 0.9826,
"step": 75000
},
{
"epoch": 1.51,
"learning_rate": 1.698e-05,
"loss": 0.9659,
"step": 75500
},
{
"epoch": 1.52,
"learning_rate": 1.696e-05,
"loss": 0.9776,
"step": 76000
},
{
"epoch": 1.53,
"learning_rate": 1.694e-05,
"loss": 0.9575,
"step": 76500
},
{
"epoch": 1.54,
"learning_rate": 1.692e-05,
"loss": 0.9614,
"step": 77000
},
{
"epoch": 1.55,
"learning_rate": 1.69e-05,
"loss": 0.9275,
"step": 77500
},
{
"epoch": 1.56,
"learning_rate": 1.688e-05,
"loss": 0.9523,
"step": 78000
},
{
"epoch": 1.57,
"learning_rate": 1.686e-05,
"loss": 0.9806,
"step": 78500
},
{
"epoch": 1.58,
"learning_rate": 1.684e-05,
"loss": 0.9169,
"step": 79000
},
{
"epoch": 1.59,
"learning_rate": 1.682e-05,
"loss": 0.9539,
"step": 79500
},
{
"epoch": 1.6,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.9378,
"step": 80000
},
{
"epoch": 1.61,
"learning_rate": 1.6780000000000002e-05,
"loss": 0.9668,
"step": 80500
},
{
"epoch": 1.62,
"learning_rate": 1.6760000000000002e-05,
"loss": 0.9496,
"step": 81000
},
{
"epoch": 1.63,
"learning_rate": 1.6740000000000002e-05,
"loss": 0.9712,
"step": 81500
},
{
"epoch": 1.64,
"learning_rate": 1.672e-05,
"loss": 0.9617,
"step": 82000
},
{
"epoch": 1.65,
"learning_rate": 1.67e-05,
"loss": 0.9189,
"step": 82500
},
{
"epoch": 1.66,
"learning_rate": 1.668e-05,
"loss": 0.9294,
"step": 83000
},
{
"epoch": 1.67,
"learning_rate": 1.666e-05,
"loss": 0.9742,
"step": 83500
},
{
"epoch": 1.68,
"learning_rate": 1.664e-05,
"loss": 0.9528,
"step": 84000
},
{
"epoch": 1.69,
"learning_rate": 1.662e-05,
"loss": 0.9349,
"step": 84500
},
{
"epoch": 1.7,
"learning_rate": 1.66e-05,
"loss": 0.981,
"step": 85000
},
{
"epoch": 1.71,
"learning_rate": 1.658e-05,
"loss": 0.9718,
"step": 85500
},
{
"epoch": 1.72,
"learning_rate": 1.656e-05,
"loss": 0.9832,
"step": 86000
},
{
"epoch": 1.73,
"learning_rate": 1.654e-05,
"loss": 0.9193,
"step": 86500
},
{
"epoch": 1.74,
"learning_rate": 1.652e-05,
"loss": 0.9687,
"step": 87000
},
{
"epoch": 1.75,
"learning_rate": 1.65e-05,
"loss": 0.9637,
"step": 87500
},
{
"epoch": 1.76,
"learning_rate": 1.648e-05,
"loss": 0.978,
"step": 88000
},
{
"epoch": 1.77,
"learning_rate": 1.646e-05,
"loss": 0.9461,
"step": 88500
},
{
"epoch": 1.78,
"learning_rate": 1.6440000000000002e-05,
"loss": 0.9461,
"step": 89000
},
{
"epoch": 1.79,
"learning_rate": 1.6420000000000002e-05,
"loss": 0.9343,
"step": 89500
},
{
"epoch": 1.8,
"learning_rate": 1.64e-05,
"loss": 0.9584,
"step": 90000
},
{
"epoch": 1.81,
"learning_rate": 1.638e-05,
"loss": 0.9405,
"step": 90500
},
{
"epoch": 1.82,
"learning_rate": 1.636e-05,
"loss": 0.9369,
"step": 91000
},
{
"epoch": 1.83,
"learning_rate": 1.634e-05,
"loss": 0.9672,
"step": 91500
},
{
"epoch": 1.84,
"learning_rate": 1.632e-05,
"loss": 0.9436,
"step": 92000
},
{
"epoch": 1.85,
"learning_rate": 1.63e-05,
"loss": 0.9468,
"step": 92500
},
{
"epoch": 1.86,
"learning_rate": 1.628e-05,
"loss": 0.928,
"step": 93000
},
{
"epoch": 1.87,
"learning_rate": 1.626e-05,
"loss": 0.9393,
"step": 93500
},
{
"epoch": 1.88,
"learning_rate": 1.6240000000000004e-05,
"loss": 0.9574,
"step": 94000
},
{
"epoch": 1.89,
"learning_rate": 1.6220000000000004e-05,
"loss": 0.9465,
"step": 94500
},
{
"epoch": 1.9,
"learning_rate": 1.62e-05,
"loss": 0.958,
"step": 95000
},
{
"epoch": 1.91,
"learning_rate": 1.618e-05,
"loss": 0.9365,
"step": 95500
},
{
"epoch": 1.92,
"learning_rate": 1.616e-05,
"loss": 0.9451,
"step": 96000
},
{
"epoch": 1.93,
"learning_rate": 1.614e-05,
"loss": 0.9642,
"step": 96500
},
{
"epoch": 1.94,
"learning_rate": 1.612e-05,
"loss": 0.9629,
"step": 97000
},
{
"epoch": 1.95,
"learning_rate": 1.6100000000000002e-05,
"loss": 0.9441,
"step": 97500
},
{
"epoch": 1.96,
"learning_rate": 1.6080000000000002e-05,
"loss": 0.9367,
"step": 98000
},
{
"epoch": 1.97,
"learning_rate": 1.6060000000000002e-05,
"loss": 0.9527,
"step": 98500
},
{
"epoch": 1.98,
"learning_rate": 1.6040000000000002e-05,
"loss": 0.9322,
"step": 99000
},
{
"epoch": 1.99,
"learning_rate": 1.6020000000000002e-05,
"loss": 0.9424,
"step": 99500
},
{
"epoch": 2.0,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.9206,
"step": 100000
},
{
"epoch": 2.0,
"eval_loss": 1.0401068925857544,
"eval_runtime": 38.2298,
"eval_samples_per_second": 220.744,
"step": 100000
},
{
"epoch": 2.01,
"learning_rate": 1.5980000000000003e-05,
"loss": 0.9206,
"step": 100500
},
{
"epoch": 2.02,
"learning_rate": 1.5960000000000003e-05,
"loss": 0.9442,
"step": 101000
},
{
"epoch": 2.03,
"learning_rate": 1.5940000000000003e-05,
"loss": 0.9081,
"step": 101500
},
{
"epoch": 2.04,
"learning_rate": 1.5920000000000003e-05,
"loss": 0.9236,
"step": 102000
},
{
"epoch": 2.05,
"learning_rate": 1.5900000000000004e-05,
"loss": 0.9012,
"step": 102500
},
{
"epoch": 2.06,
"learning_rate": 1.588e-05,
"loss": 0.9594,
"step": 103000
},
{
"epoch": 2.07,
"learning_rate": 1.586e-05,
"loss": 0.9216,
"step": 103500
},
{
"epoch": 2.08,
"learning_rate": 1.584e-05,
"loss": 0.9125,
"step": 104000
},
{
"epoch": 2.09,
"learning_rate": 1.582e-05,
"loss": 0.9117,
"step": 104500
},
{
"epoch": 2.1,
"learning_rate": 1.58e-05,
"loss": 0.9444,
"step": 105000
},
{
"epoch": 2.11,
"learning_rate": 1.578e-05,
"loss": 0.9321,
"step": 105500
},
{
"epoch": 2.12,
"learning_rate": 1.576e-05,
"loss": 0.9515,
"step": 106000
},
{
"epoch": 2.13,
"learning_rate": 1.5740000000000002e-05,
"loss": 0.9496,
"step": 106500
},
{
"epoch": 2.14,
"learning_rate": 1.5720000000000002e-05,
"loss": 0.9315,
"step": 107000
},
{
"epoch": 2.15,
"learning_rate": 1.5700000000000002e-05,
"loss": 0.9478,
"step": 107500
},
{
"epoch": 2.16,
"learning_rate": 1.5680000000000002e-05,
"loss": 0.938,
"step": 108000
},
{
"epoch": 2.17,
"learning_rate": 1.5660000000000003e-05,
"loss": 0.8838,
"step": 108500
},
{
"epoch": 2.18,
"learning_rate": 1.5640000000000003e-05,
"loss": 0.9082,
"step": 109000
},
{
"epoch": 2.19,
"learning_rate": 1.5620000000000003e-05,
"loss": 0.9313,
"step": 109500
},
{
"epoch": 2.2,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.917,
"step": 110000
},
{
"epoch": 2.21,
"learning_rate": 1.5580000000000003e-05,
"loss": 0.9334,
"step": 110500
},
{
"epoch": 2.22,
"learning_rate": 1.556e-05,
"loss": 0.9425,
"step": 111000
},
{
"epoch": 2.23,
"learning_rate": 1.554e-05,
"loss": 0.9317,
"step": 111500
},
{
"epoch": 2.24,
"learning_rate": 1.552e-05,
"loss": 0.9415,
"step": 112000
},
{
"epoch": 2.25,
"learning_rate": 1.55e-05,
"loss": 0.9138,
"step": 112500
},
{
"epoch": 2.26,
"learning_rate": 1.548e-05,
"loss": 0.9132,
"step": 113000
},
{
"epoch": 2.27,
"learning_rate": 1.546e-05,
"loss": 0.9338,
"step": 113500
},
{
"epoch": 2.28,
"learning_rate": 1.544e-05,
"loss": 0.9474,
"step": 114000
},
{
"epoch": 2.29,
"learning_rate": 1.542e-05,
"loss": 0.9141,
"step": 114500
},
{
"epoch": 2.3,
"learning_rate": 1.54e-05,
"loss": 0.9,
"step": 115000
},
{
"epoch": 2.31,
"learning_rate": 1.5380000000000002e-05,
"loss": 0.9178,
"step": 115500
},
{
"epoch": 2.32,
"learning_rate": 1.5360000000000002e-05,
"loss": 0.9222,
"step": 116000
},
{
"epoch": 2.33,
"learning_rate": 1.5340000000000002e-05,
"loss": 0.8756,
"step": 116500
},
{
"epoch": 2.34,
"learning_rate": 1.5320000000000002e-05,
"loss": 0.9079,
"step": 117000
},
{
"epoch": 2.35,
"learning_rate": 1.5300000000000003e-05,
"loss": 0.9154,
"step": 117500
},
{
"epoch": 2.36,
"learning_rate": 1.5280000000000003e-05,
"loss": 0.9191,
"step": 118000
},
{
"epoch": 2.37,
"learning_rate": 1.5260000000000003e-05,
"loss": 0.9101,
"step": 118500
},
{
"epoch": 2.38,
"learning_rate": 1.5240000000000001e-05,
"loss": 0.9268,
"step": 119000
},
{
"epoch": 2.39,
"learning_rate": 1.5220000000000002e-05,
"loss": 0.9455,
"step": 119500
},
{
"epoch": 2.4,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.8976,
"step": 120000
},
{
"epoch": 2.41,
"learning_rate": 1.5180000000000002e-05,
"loss": 0.9389,
"step": 120500
},
{
"epoch": 2.42,
"learning_rate": 1.516e-05,
"loss": 0.9626,
"step": 121000
},
{
"epoch": 2.43,
"learning_rate": 1.514e-05,
"loss": 0.9094,
"step": 121500
},
{
"epoch": 2.44,
"learning_rate": 1.5120000000000001e-05,
"loss": 0.9505,
"step": 122000
},
{
"epoch": 2.45,
"learning_rate": 1.5100000000000001e-05,
"loss": 0.941,
"step": 122500
},
{
"epoch": 2.46,
"learning_rate": 1.5080000000000001e-05,
"loss": 0.8881,
"step": 123000
},
{
"epoch": 2.47,
"learning_rate": 1.5060000000000001e-05,
"loss": 0.8981,
"step": 123500
},
{
"epoch": 2.48,
"learning_rate": 1.5040000000000002e-05,
"loss": 0.935,
"step": 124000
},
{
"epoch": 2.49,
"learning_rate": 1.5020000000000002e-05,
"loss": 0.9324,
"step": 124500
},
{
"epoch": 2.5,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.9057,
"step": 125000
},
{
"epoch": 2.51,
"learning_rate": 1.498e-05,
"loss": 0.9071,
"step": 125500
},
{
"epoch": 2.52,
"learning_rate": 1.496e-05,
"loss": 0.9294,
"step": 126000
},
{
"epoch": 2.53,
"learning_rate": 1.4940000000000001e-05,
"loss": 0.9223,
"step": 126500
},
{
"epoch": 2.54,
"learning_rate": 1.4920000000000001e-05,
"loss": 0.9375,
"step": 127000
},
{
"epoch": 2.55,
"learning_rate": 1.4900000000000001e-05,
"loss": 0.9598,
"step": 127500
},
{
"epoch": 2.56,
"learning_rate": 1.4880000000000002e-05,
"loss": 0.9026,
"step": 128000
},
{
"epoch": 2.57,
"learning_rate": 1.4860000000000002e-05,
"loss": 0.9126,
"step": 128500
},
{
"epoch": 2.58,
"learning_rate": 1.4840000000000002e-05,
"loss": 0.9249,
"step": 129000
},
{
"epoch": 2.59,
"learning_rate": 1.482e-05,
"loss": 0.9315,
"step": 129500
},
{
"epoch": 2.6,
"learning_rate": 1.48e-05,
"loss": 0.9222,
"step": 130000
},
{
"epoch": 2.61,
"learning_rate": 1.478e-05,
"loss": 0.9345,
"step": 130500
},
{
"epoch": 2.62,
"learning_rate": 1.4760000000000001e-05,
"loss": 0.907,
"step": 131000
},
{
"epoch": 2.63,
"learning_rate": 1.4740000000000001e-05,
"loss": 0.9335,
"step": 131500
},
{
"epoch": 2.64,
"learning_rate": 1.4720000000000001e-05,
"loss": 0.9016,
"step": 132000
},
{
"epoch": 2.65,
"learning_rate": 1.4700000000000002e-05,
"loss": 0.9286,
"step": 132500
},
{
"epoch": 2.66,
"learning_rate": 1.4680000000000002e-05,
"loss": 0.9309,
"step": 133000
},
{
"epoch": 2.67,
"learning_rate": 1.466e-05,
"loss": 0.8985,
"step": 133500
},
{
"epoch": 2.68,
"learning_rate": 1.464e-05,
"loss": 0.931,
"step": 134000
},
{
"epoch": 2.69,
"learning_rate": 1.462e-05,
"loss": 0.8933,
"step": 134500
},
{
"epoch": 2.7,
"learning_rate": 1.46e-05,
"loss": 0.9064,
"step": 135000
},
{
"epoch": 2.71,
"learning_rate": 1.4580000000000001e-05,
"loss": 0.8963,
"step": 135500
},
{
"epoch": 2.72,
"learning_rate": 1.4560000000000001e-05,
"loss": 0.9055,
"step": 136000
},
{
"epoch": 2.73,
"learning_rate": 1.4540000000000001e-05,
"loss": 0.8748,
"step": 136500
},
{
"epoch": 2.74,
"learning_rate": 1.4520000000000002e-05,
"loss": 0.9339,
"step": 137000
},
{
"epoch": 2.75,
"learning_rate": 1.45e-05,
"loss": 0.9336,
"step": 137500
},
{
"epoch": 2.76,
"learning_rate": 1.448e-05,
"loss": 0.9413,
"step": 138000
},
{
"epoch": 2.77,
"learning_rate": 1.446e-05,
"loss": 0.8967,
"step": 138500
},
{
"epoch": 2.78,
"learning_rate": 1.444e-05,
"loss": 0.9144,
"step": 139000
},
{
"epoch": 2.79,
"learning_rate": 1.4420000000000001e-05,
"loss": 0.9165,
"step": 139500
},
{
"epoch": 2.8,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.9088,
"step": 140000
},
{
"epoch": 2.81,
"learning_rate": 1.4380000000000001e-05,
"loss": 0.9127,
"step": 140500
},
{
"epoch": 2.82,
"learning_rate": 1.4360000000000001e-05,
"loss": 0.9226,
"step": 141000
},
{
"epoch": 2.83,
"learning_rate": 1.434e-05,
"loss": 0.9374,
"step": 141500
},
{
"epoch": 2.84,
"learning_rate": 1.432e-05,
"loss": 0.9111,
"step": 142000
},
{
"epoch": 2.85,
"learning_rate": 1.43e-05,
"loss": 0.9213,
"step": 142500
},
{
"epoch": 2.86,
"learning_rate": 1.428e-05,
"loss": 0.9115,
"step": 143000
},
{
"epoch": 2.87,
"learning_rate": 1.426e-05,
"loss": 0.9207,
"step": 143500
},
{
"epoch": 2.88,
"learning_rate": 1.4240000000000001e-05,
"loss": 0.9162,
"step": 144000
},
{
"epoch": 2.89,
"learning_rate": 1.4220000000000001e-05,
"loss": 0.9338,
"step": 144500
},
{
"epoch": 2.9,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.9201,
"step": 145000
},
{
"epoch": 2.91,
"learning_rate": 1.418e-05,
"loss": 0.8941,
"step": 145500
},
{
"epoch": 2.92,
"learning_rate": 1.416e-05,
"loss": 0.8942,
"step": 146000
},
{
"epoch": 2.93,
"learning_rate": 1.414e-05,
"loss": 0.9272,
"step": 146500
},
{
"epoch": 2.94,
"learning_rate": 1.412e-05,
"loss": 0.9401,
"step": 147000
},
{
"epoch": 2.95,
"learning_rate": 1.41e-05,
"loss": 0.9248,
"step": 147500
},
{
"epoch": 2.96,
"learning_rate": 1.408e-05,
"loss": 0.9509,
"step": 148000
},
{
"epoch": 2.97,
"learning_rate": 1.4060000000000001e-05,
"loss": 0.9258,
"step": 148500
},
{
"epoch": 2.98,
"learning_rate": 1.4040000000000001e-05,
"loss": 0.914,
"step": 149000
},
{
"epoch": 2.99,
"learning_rate": 1.402e-05,
"loss": 0.9115,
"step": 149500
},
{
"epoch": 3.0,
"learning_rate": 1.4e-05,
"loss": 0.9401,
"step": 150000
},
{
"epoch": 3.0,
"eval_loss": 1.0360552072525024,
"eval_runtime": 38.0361,
"eval_samples_per_second": 221.868,
"step": 150000
},
{
"epoch": 3.01,
"learning_rate": 1.398e-05,
"loss": 0.8871,
"step": 150500
},
{
"epoch": 3.02,
"learning_rate": 1.396e-05,
"loss": 0.885,
"step": 151000
},
{
"epoch": 3.03,
"learning_rate": 1.394e-05,
"loss": 0.9214,
"step": 151500
},
{
"epoch": 3.04,
"learning_rate": 1.392e-05,
"loss": 0.9042,
"step": 152000
},
{
"epoch": 3.05,
"learning_rate": 1.39e-05,
"loss": 0.8641,
"step": 152500
},
{
"epoch": 3.06,
"learning_rate": 1.3880000000000001e-05,
"loss": 0.8995,
"step": 153000
},
{
"epoch": 3.07,
"learning_rate": 1.386e-05,
"loss": 0.9037,
"step": 153500
},
{
"epoch": 3.08,
"learning_rate": 1.384e-05,
"loss": 0.9184,
"step": 154000
},
{
"epoch": 3.09,
"learning_rate": 1.382e-05,
"loss": 0.9066,
"step": 154500
},
{
"epoch": 3.1,
"learning_rate": 1.38e-05,
"loss": 0.8938,
"step": 155000
},
{
"epoch": 3.11,
"learning_rate": 1.378e-05,
"loss": 0.9076,
"step": 155500
},
{
"epoch": 3.12,
"learning_rate": 1.376e-05,
"loss": 0.8916,
"step": 156000
},
{
"epoch": 3.13,
"learning_rate": 1.3740000000000002e-05,
"loss": 0.8846,
"step": 156500
},
{
"epoch": 3.14,
"learning_rate": 1.3720000000000002e-05,
"loss": 0.887,
"step": 157000
},
{
"epoch": 3.15,
"learning_rate": 1.3700000000000003e-05,
"loss": 0.8879,
"step": 157500
},
{
"epoch": 3.16,
"learning_rate": 1.3680000000000003e-05,
"loss": 0.9067,
"step": 158000
},
{
"epoch": 3.17,
"learning_rate": 1.3660000000000001e-05,
"loss": 0.9291,
"step": 158500
},
{
"epoch": 3.18,
"learning_rate": 1.3640000000000002e-05,
"loss": 0.9288,
"step": 159000
},
{
"epoch": 3.19,
"learning_rate": 1.3620000000000002e-05,
"loss": 0.869,
"step": 159500
},
{
"epoch": 3.2,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.9057,
"step": 160000
},
{
"epoch": 3.21,
"learning_rate": 1.3580000000000002e-05,
"loss": 0.8712,
"step": 160500
},
{
"epoch": 3.22,
"learning_rate": 1.3560000000000002e-05,
"loss": 0.9002,
"step": 161000
},
{
"epoch": 3.23,
"learning_rate": 1.3540000000000003e-05,
"loss": 0.8941,
"step": 161500
},
{
"epoch": 3.24,
"learning_rate": 1.3520000000000003e-05,
"loss": 0.9078,
"step": 162000
},
{
"epoch": 3.25,
"learning_rate": 1.3500000000000001e-05,
"loss": 0.9013,
"step": 162500
},
{
"epoch": 3.26,
"learning_rate": 1.3480000000000001e-05,
"loss": 0.9113,
"step": 163000
},
{
"epoch": 3.27,
"learning_rate": 1.3460000000000002e-05,
"loss": 0.9183,
"step": 163500
},
{
"epoch": 3.28,
"learning_rate": 1.3440000000000002e-05,
"loss": 0.8761,
"step": 164000
},
{
"epoch": 3.29,
"learning_rate": 1.3420000000000002e-05,
"loss": 0.9013,
"step": 164500
},
{
"epoch": 3.3,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.9201,
"step": 165000
},
{
"epoch": 3.31,
"learning_rate": 1.3380000000000002e-05,
"loss": 0.9147,
"step": 165500
},
{
"epoch": 3.32,
"learning_rate": 1.3360000000000003e-05,
"loss": 0.8999,
"step": 166000
},
{
"epoch": 3.33,
"learning_rate": 1.3340000000000001e-05,
"loss": 0.8977,
"step": 166500
},
{
"epoch": 3.34,
"learning_rate": 1.3320000000000001e-05,
"loss": 0.9165,
"step": 167000
},
{
"epoch": 3.35,
"learning_rate": 1.3300000000000001e-05,
"loss": 0.88,
"step": 167500
},
{
"epoch": 3.36,
"learning_rate": 1.3280000000000002e-05,
"loss": 0.8984,
"step": 168000
},
{
"epoch": 3.37,
"learning_rate": 1.3260000000000002e-05,
"loss": 0.9192,
"step": 168500
},
{
"epoch": 3.38,
"learning_rate": 1.3240000000000002e-05,
"loss": 0.9002,
"step": 169000
},
{
"epoch": 3.39,
"learning_rate": 1.3220000000000002e-05,
"loss": 0.9063,
"step": 169500
},
{
"epoch": 3.4,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.8966,
"step": 170000
},
{
"epoch": 3.41,
"learning_rate": 1.3180000000000001e-05,
"loss": 0.8963,
"step": 170500
},
{
"epoch": 3.42,
"learning_rate": 1.3160000000000001e-05,
"loss": 0.9023,
"step": 171000
},
{
"epoch": 3.43,
"learning_rate": 1.3140000000000001e-05,
"loss": 0.9139,
"step": 171500
},
{
"epoch": 3.44,
"learning_rate": 1.3120000000000001e-05,
"loss": 0.8753,
"step": 172000
},
{
"epoch": 3.45,
"learning_rate": 1.3100000000000002e-05,
"loss": 0.9033,
"step": 172500
},
{
"epoch": 3.46,
"learning_rate": 1.3080000000000002e-05,
"loss": 0.8899,
"step": 173000
},
{
"epoch": 3.47,
"learning_rate": 1.3060000000000002e-05,
"loss": 0.9349,
"step": 173500
},
{
"epoch": 3.48,
"learning_rate": 1.3040000000000002e-05,
"loss": 0.9069,
"step": 174000
},
{
"epoch": 3.49,
"learning_rate": 1.302e-05,
"loss": 0.8968,
"step": 174500
},
{
"epoch": 3.5,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.8852,
"step": 175000
},
{
"epoch": 3.51,
"learning_rate": 1.2980000000000001e-05,
"loss": 0.8884,
"step": 175500
},
{
"epoch": 3.52,
"learning_rate": 1.2960000000000001e-05,
"loss": 0.9019,
"step": 176000
},
{
"epoch": 3.53,
"learning_rate": 1.2940000000000001e-05,
"loss": 0.8991,
"step": 176500
},
{
"epoch": 3.54,
"learning_rate": 1.2920000000000002e-05,
"loss": 0.8742,
"step": 177000
},
{
"epoch": 3.55,
"learning_rate": 1.2900000000000002e-05,
"loss": 0.9328,
"step": 177500
},
{
"epoch": 3.56,
"learning_rate": 1.2880000000000002e-05,
"loss": 0.9012,
"step": 178000
},
{
"epoch": 3.57,
"learning_rate": 1.286e-05,
"loss": 0.9047,
"step": 178500
},
{
"epoch": 3.58,
"learning_rate": 1.284e-05,
"loss": 0.8978,
"step": 179000
},
{
"epoch": 3.59,
"learning_rate": 1.2820000000000001e-05,
"loss": 0.8974,
"step": 179500
},
{
"epoch": 3.6,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.8864,
"step": 180000
},
{
"epoch": 3.61,
"learning_rate": 1.2780000000000001e-05,
"loss": 0.883,
"step": 180500
},
{
"epoch": 3.62,
"learning_rate": 1.2760000000000001e-05,
"loss": 0.9191,
"step": 181000
},
{
"epoch": 3.63,
"learning_rate": 1.2740000000000002e-05,
"loss": 0.9185,
"step": 181500
},
{
"epoch": 3.64,
"learning_rate": 1.2720000000000002e-05,
"loss": 0.8723,
"step": 182000
},
{
"epoch": 3.65,
"learning_rate": 1.27e-05,
"loss": 0.8941,
"step": 182500
},
{
"epoch": 3.66,
"learning_rate": 1.268e-05,
"loss": 0.8825,
"step": 183000
},
{
"epoch": 3.67,
"learning_rate": 1.266e-05,
"loss": 0.8716,
"step": 183500
},
{
"epoch": 3.68,
"learning_rate": 1.2640000000000001e-05,
"loss": 0.8752,
"step": 184000
},
{
"epoch": 3.69,
"learning_rate": 1.2620000000000001e-05,
"loss": 0.9101,
"step": 184500
},
{
"epoch": 3.7,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.9017,
"step": 185000
},
{
"epoch": 3.71,
"learning_rate": 1.2580000000000002e-05,
"loss": 0.9123,
"step": 185500
},
{
"epoch": 3.72,
"learning_rate": 1.2560000000000002e-05,
"loss": 0.9461,
"step": 186000
},
{
"epoch": 3.73,
"learning_rate": 1.254e-05,
"loss": 0.915,
"step": 186500
},
{
"epoch": 3.74,
"learning_rate": 1.252e-05,
"loss": 0.8812,
"step": 187000
},
{
"epoch": 3.75,
"learning_rate": 1.25e-05,
"loss": 0.9088,
"step": 187500
},
{
"epoch": 3.76,
"learning_rate": 1.248e-05,
"loss": 0.9244,
"step": 188000
},
{
"epoch": 3.77,
"learning_rate": 1.2460000000000001e-05,
"loss": 0.8963,
"step": 188500
},
{
"epoch": 3.78,
"learning_rate": 1.2440000000000001e-05,
"loss": 0.88,
"step": 189000
},
{
"epoch": 3.79,
"learning_rate": 1.2420000000000001e-05,
"loss": 0.8912,
"step": 189500
},
{
"epoch": 3.8,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.8959,
"step": 190000
},
{
"epoch": 3.81,
"learning_rate": 1.2380000000000002e-05,
"loss": 0.8935,
"step": 190500
},
{
"epoch": 3.82,
"learning_rate": 1.236e-05,
"loss": 0.8894,
"step": 191000
},
{
"epoch": 3.83,
"learning_rate": 1.234e-05,
"loss": 0.8805,
"step": 191500
},
{
"epoch": 3.84,
"learning_rate": 1.232e-05,
"loss": 0.9075,
"step": 192000
},
{
"epoch": 3.85,
"learning_rate": 1.23e-05,
"loss": 0.8917,
"step": 192500
},
{
"epoch": 3.86,
"learning_rate": 1.2280000000000001e-05,
"loss": 0.8614,
"step": 193000
},
{
"epoch": 3.87,
"learning_rate": 1.2260000000000001e-05,
"loss": 0.8729,
"step": 193500
},
{
"epoch": 3.88,
"learning_rate": 1.2240000000000001e-05,
"loss": 0.8799,
"step": 194000
},
{
"epoch": 3.89,
"learning_rate": 1.2220000000000002e-05,
"loss": 0.9237,
"step": 194500
},
{
"epoch": 3.9,
"learning_rate": 1.22e-05,
"loss": 0.9123,
"step": 195000
},
{
"epoch": 3.91,
"learning_rate": 1.218e-05,
"loss": 0.9212,
"step": 195500
},
{
"epoch": 3.92,
"learning_rate": 1.216e-05,
"loss": 0.8751,
"step": 196000
},
{
"epoch": 3.93,
"learning_rate": 1.214e-05,
"loss": 0.8942,
"step": 196500
},
{
"epoch": 3.94,
"learning_rate": 1.2120000000000001e-05,
"loss": 0.8926,
"step": 197000
},
{
"epoch": 3.95,
"learning_rate": 1.2100000000000001e-05,
"loss": 0.8766,
"step": 197500
},
{
"epoch": 3.96,
"learning_rate": 1.2080000000000001e-05,
"loss": 0.8951,
"step": 198000
},
{
"epoch": 3.97,
"learning_rate": 1.2060000000000001e-05,
"loss": 0.9063,
"step": 198500
},
{
"epoch": 3.98,
"learning_rate": 1.204e-05,
"loss": 0.8995,
"step": 199000
},
{
"epoch": 3.99,
"learning_rate": 1.202e-05,
"loss": 0.8949,
"step": 199500
},
{
"epoch": 4.0,
"learning_rate": 1.2e-05,
"loss": 0.9091,
"step": 200000
},
{
"epoch": 4.0,
"eval_loss": 1.0091001987457275,
"eval_runtime": 34.9666,
"eval_samples_per_second": 241.345,
"step": 200000
},
{
"epoch": 4.01,
"learning_rate": 1.198e-05,
"loss": 0.8757,
"step": 200500
},
{
"epoch": 4.02,
"learning_rate": 1.196e-05,
"loss": 0.8703,
"step": 201000
},
{
"epoch": 4.03,
"learning_rate": 1.1940000000000001e-05,
"loss": 0.9002,
"step": 201500
},
{
"epoch": 4.04,
"learning_rate": 1.1920000000000001e-05,
"loss": 0.8746,
"step": 202000
},
{
"epoch": 4.05,
"learning_rate": 1.1900000000000001e-05,
"loss": 0.8543,
"step": 202500
},
{
"epoch": 4.06,
"learning_rate": 1.188e-05,
"loss": 0.8602,
"step": 203000
},
{
"epoch": 4.07,
"learning_rate": 1.186e-05,
"loss": 0.8668,
"step": 203500
},
{
"epoch": 4.08,
"learning_rate": 1.184e-05,
"loss": 0.8917,
"step": 204000
},
{
"epoch": 4.09,
"learning_rate": 1.182e-05,
"loss": 0.8905,
"step": 204500
},
{
"epoch": 4.1,
"learning_rate": 1.18e-05,
"loss": 0.8664,
"step": 205000
},
{
"epoch": 4.11,
"learning_rate": 1.178e-05,
"loss": 0.8751,
"step": 205500
},
{
"epoch": 4.12,
"learning_rate": 1.1760000000000001e-05,
"loss": 0.8781,
"step": 206000
},
{
"epoch": 4.13,
"learning_rate": 1.1740000000000001e-05,
"loss": 0.8737,
"step": 206500
},
{
"epoch": 4.14,
"learning_rate": 1.172e-05,
"loss": 0.8616,
"step": 207000
},
{
"epoch": 4.15,
"learning_rate": 1.17e-05,
"loss": 0.8776,
"step": 207500
},
{
"epoch": 4.16,
"learning_rate": 1.168e-05,
"loss": 0.9032,
"step": 208000
},
{
"epoch": 4.17,
"learning_rate": 1.166e-05,
"loss": 0.8972,
"step": 208500
},
{
"epoch": 4.18,
"learning_rate": 1.164e-05,
"loss": 0.89,
"step": 209000
},
{
"epoch": 4.19,
"learning_rate": 1.162e-05,
"loss": 0.863,
"step": 209500
},
{
"epoch": 4.2,
"learning_rate": 1.16e-05,
"loss": 0.8609,
"step": 210000
},
{
"epoch": 4.21,
"learning_rate": 1.1580000000000001e-05,
"loss": 0.8667,
"step": 210500
},
{
"epoch": 4.22,
"learning_rate": 1.156e-05,
"loss": 0.8581,
"step": 211000
},
{
"epoch": 4.23,
"learning_rate": 1.154e-05,
"loss": 0.8668,
"step": 211500
},
{
"epoch": 4.24,
"learning_rate": 1.152e-05,
"loss": 0.9037,
"step": 212000
},
{
"epoch": 4.25,
"learning_rate": 1.15e-05,
"loss": 0.9107,
"step": 212500
},
{
"epoch": 4.26,
"learning_rate": 1.148e-05,
"loss": 0.8756,
"step": 213000
},
{
"epoch": 4.27,
"learning_rate": 1.146e-05,
"loss": 0.8658,
"step": 213500
},
{
"epoch": 4.28,
"learning_rate": 1.144e-05,
"loss": 0.8647,
"step": 214000
},
{
"epoch": 4.29,
"learning_rate": 1.142e-05,
"loss": 0.9025,
"step": 214500
},
{
"epoch": 4.3,
"learning_rate": 1.14e-05,
"loss": 0.8775,
"step": 215000
},
{
"epoch": 4.31,
"learning_rate": 1.138e-05,
"loss": 0.8759,
"step": 215500
},
{
"epoch": 4.32,
"learning_rate": 1.136e-05,
"loss": 0.8681,
"step": 216000
},
{
"epoch": 4.33,
"learning_rate": 1.134e-05,
"loss": 0.8751,
"step": 216500
},
{
"epoch": 4.34,
"learning_rate": 1.132e-05,
"loss": 0.9005,
"step": 217000
},
{
"epoch": 4.35,
"learning_rate": 1.13e-05,
"loss": 0.8791,
"step": 217500
},
{
"epoch": 4.36,
"learning_rate": 1.128e-05,
"loss": 0.8595,
"step": 218000
},
{
"epoch": 4.37,
"learning_rate": 1.126e-05,
"loss": 0.905,
"step": 218500
},
{
"epoch": 4.38,
"learning_rate": 1.1240000000000002e-05,
"loss": 0.9036,
"step": 219000
},
{
"epoch": 4.39,
"learning_rate": 1.1220000000000003e-05,
"loss": 0.8793,
"step": 219500
},
{
"epoch": 4.4,
"learning_rate": 1.1200000000000001e-05,
"loss": 0.8686,
"step": 220000
},
{
"epoch": 4.41,
"learning_rate": 1.1180000000000001e-05,
"loss": 0.8397,
"step": 220500
},
{
"epoch": 4.42,
"learning_rate": 1.1160000000000002e-05,
"loss": 0.8454,
"step": 221000
},
{
"epoch": 4.43,
"learning_rate": 1.1140000000000002e-05,
"loss": 0.8576,
"step": 221500
},
{
"epoch": 4.44,
"learning_rate": 1.1120000000000002e-05,
"loss": 0.8645,
"step": 222000
},
{
"epoch": 4.45,
"learning_rate": 1.1100000000000002e-05,
"loss": 0.882,
"step": 222500
},
{
"epoch": 4.46,
"learning_rate": 1.1080000000000002e-05,
"loss": 0.8672,
"step": 223000
},
{
"epoch": 4.47,
"learning_rate": 1.1060000000000003e-05,
"loss": 0.8772,
"step": 223500
},
{
"epoch": 4.48,
"learning_rate": 1.1040000000000001e-05,
"loss": 0.8743,
"step": 224000
},
{
"epoch": 4.49,
"learning_rate": 1.1020000000000001e-05,
"loss": 0.8809,
"step": 224500
},
{
"epoch": 4.5,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.865,
"step": 225000
},
{
"epoch": 4.51,
"learning_rate": 1.0980000000000002e-05,
"loss": 0.8971,
"step": 225500
},
{
"epoch": 4.52,
"learning_rate": 1.0960000000000002e-05,
"loss": 0.8752,
"step": 226000
},
{
"epoch": 4.53,
"learning_rate": 1.0940000000000002e-05,
"loss": 0.8505,
"step": 226500
},
{
"epoch": 4.54,
"learning_rate": 1.0920000000000002e-05,
"loss": 0.845,
"step": 227000
},
{
"epoch": 4.55,
"learning_rate": 1.0900000000000002e-05,
"loss": 0.873,
"step": 227500
},
{
"epoch": 4.56,
"learning_rate": 1.0880000000000001e-05,
"loss": 0.914,
"step": 228000
},
{
"epoch": 4.57,
"learning_rate": 1.0860000000000001e-05,
"loss": 0.8504,
"step": 228500
},
{
"epoch": 4.58,
"learning_rate": 1.0840000000000001e-05,
"loss": 0.8951,
"step": 229000
},
{
"epoch": 4.59,
"learning_rate": 1.0820000000000001e-05,
"loss": 0.8836,
"step": 229500
},
{
"epoch": 4.6,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.8557,
"step": 230000
},
{
"epoch": 4.61,
"learning_rate": 1.0780000000000002e-05,
"loss": 0.8748,
"step": 230500
},
{
"epoch": 4.62,
"learning_rate": 1.0760000000000002e-05,
"loss": 0.8803,
"step": 231000
},
{
"epoch": 4.63,
"learning_rate": 1.0740000000000002e-05,
"loss": 0.9195,
"step": 231500
},
{
"epoch": 4.64,
"learning_rate": 1.072e-05,
"loss": 0.8697,
"step": 232000
},
{
"epoch": 4.65,
"learning_rate": 1.0700000000000001e-05,
"loss": 0.8872,
"step": 232500
},
{
"epoch": 4.66,
"learning_rate": 1.0680000000000001e-05,
"loss": 0.8618,
"step": 233000
},
{
"epoch": 4.67,
"learning_rate": 1.0660000000000001e-05,
"loss": 0.8387,
"step": 233500
},
{
"epoch": 4.68,
"learning_rate": 1.0640000000000001e-05,
"loss": 0.8702,
"step": 234000
},
{
"epoch": 4.69,
"learning_rate": 1.0620000000000002e-05,
"loss": 0.8726,
"step": 234500
},
{
"epoch": 4.7,
"learning_rate": 1.0600000000000002e-05,
"loss": 0.8956,
"step": 235000
},
{
"epoch": 4.71,
"learning_rate": 1.0580000000000002e-05,
"loss": 0.8636,
"step": 235500
},
{
"epoch": 4.72,
"learning_rate": 1.056e-05,
"loss": 0.8649,
"step": 236000
},
{
"epoch": 4.73,
"learning_rate": 1.054e-05,
"loss": 0.8856,
"step": 236500
},
{
"epoch": 4.74,
"learning_rate": 1.0520000000000001e-05,
"loss": 0.8849,
"step": 237000
},
{
"epoch": 4.75,
"learning_rate": 1.0500000000000001e-05,
"loss": 0.8673,
"step": 237500
},
{
"epoch": 4.76,
"learning_rate": 1.0480000000000001e-05,
"loss": 0.8557,
"step": 238000
},
{
"epoch": 4.77,
"learning_rate": 1.0460000000000001e-05,
"loss": 0.8508,
"step": 238500
},
{
"epoch": 4.78,
"learning_rate": 1.0440000000000002e-05,
"loss": 0.8562,
"step": 239000
},
{
"epoch": 4.79,
"learning_rate": 1.0420000000000002e-05,
"loss": 0.8952,
"step": 239500
},
{
"epoch": 4.8,
"learning_rate": 1.04e-05,
"loss": 0.8949,
"step": 240000
},
{
"epoch": 4.81,
"learning_rate": 1.038e-05,
"loss": 0.882,
"step": 240500
},
{
"epoch": 4.82,
"learning_rate": 1.036e-05,
"loss": 0.8593,
"step": 241000
},
{
"epoch": 4.83,
"learning_rate": 1.0340000000000001e-05,
"loss": 0.879,
"step": 241500
},
{
"epoch": 4.84,
"learning_rate": 1.0320000000000001e-05,
"loss": 0.9084,
"step": 242000
},
{
"epoch": 4.85,
"learning_rate": 1.0300000000000001e-05,
"loss": 0.871,
"step": 242500
},
{
"epoch": 4.86,
"learning_rate": 1.0280000000000002e-05,
"loss": 0.8417,
"step": 243000
},
{
"epoch": 4.87,
"learning_rate": 1.0260000000000002e-05,
"loss": 0.8789,
"step": 243500
},
{
"epoch": 4.88,
"learning_rate": 1.024e-05,
"loss": 0.8516,
"step": 244000
},
{
"epoch": 4.89,
"learning_rate": 1.022e-05,
"loss": 0.893,
"step": 244500
},
{
"epoch": 4.9,
"learning_rate": 1.02e-05,
"loss": 0.8977,
"step": 245000
},
{
"epoch": 4.91,
"learning_rate": 1.018e-05,
"loss": 0.8702,
"step": 245500
},
{
"epoch": 4.92,
"learning_rate": 1.0160000000000001e-05,
"loss": 0.8389,
"step": 246000
},
{
"epoch": 4.93,
"learning_rate": 1.0140000000000001e-05,
"loss": 0.8826,
"step": 246500
},
{
"epoch": 4.94,
"learning_rate": 1.0120000000000001e-05,
"loss": 0.8712,
"step": 247000
},
{
"epoch": 4.95,
"learning_rate": 1.0100000000000002e-05,
"loss": 0.8674,
"step": 247500
},
{
"epoch": 4.96,
"learning_rate": 1.008e-05,
"loss": 0.8871,
"step": 248000
},
{
"epoch": 4.97,
"learning_rate": 1.006e-05,
"loss": 0.8929,
"step": 248500
},
{
"epoch": 4.98,
"learning_rate": 1.004e-05,
"loss": 0.8592,
"step": 249000
},
{
"epoch": 4.99,
"learning_rate": 1.002e-05,
"loss": 0.8824,
"step": 249500
},
{
"epoch": 5.0,
"learning_rate": 1e-05,
"loss": 0.8465,
"step": 250000
},
{
"epoch": 5.0,
"eval_loss": 1.0072413682937622,
"eval_runtime": 34.9811,
"eval_samples_per_second": 241.244,
"step": 250000
}
],
"max_steps": 500000,
"num_train_epochs": 10,
"total_flos": 6.924912949067942e+16,
"trial_name": null,
"trial_params": null
}