clip-mixer-big / trainer_state.json
bilbo991's picture
End of training
fb01ed0
{
"best_metric": 1.1062601804733276,
"best_model_checkpoint": "clip-mixer-big/checkpoint-206250",
"epoch": 3.0,
"global_step": 206250,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.987878787878788e-05,
"loss": 3.2436,
"step": 500
},
{
"epoch": 0.01,
"learning_rate": 4.9757575757575756e-05,
"loss": 2.9607,
"step": 1000
},
{
"epoch": 0.02,
"learning_rate": 4.963636363636364e-05,
"loss": 2.8062,
"step": 1500
},
{
"epoch": 0.03,
"learning_rate": 4.951515151515152e-05,
"loss": 2.7162,
"step": 2000
},
{
"epoch": 0.04,
"learning_rate": 4.93939393939394e-05,
"loss": 2.6372,
"step": 2500
},
{
"epoch": 0.04,
"learning_rate": 4.9272727272727276e-05,
"loss": 2.6116,
"step": 3000
},
{
"epoch": 0.05,
"learning_rate": 4.915151515151515e-05,
"loss": 2.5324,
"step": 3500
},
{
"epoch": 0.06,
"learning_rate": 4.903030303030303e-05,
"loss": 2.5136,
"step": 4000
},
{
"epoch": 0.07,
"learning_rate": 4.890909090909091e-05,
"loss": 2.4633,
"step": 4500
},
{
"epoch": 0.07,
"learning_rate": 4.878787878787879e-05,
"loss": 2.4195,
"step": 5000
},
{
"epoch": 0.08,
"learning_rate": 4.866666666666667e-05,
"loss": 2.3985,
"step": 5500
},
{
"epoch": 0.09,
"learning_rate": 4.854545454545455e-05,
"loss": 2.3739,
"step": 6000
},
{
"epoch": 0.09,
"learning_rate": 4.8424242424242426e-05,
"loss": 2.3513,
"step": 6500
},
{
"epoch": 0.1,
"learning_rate": 4.83030303030303e-05,
"loss": 2.3144,
"step": 7000
},
{
"epoch": 0.11,
"learning_rate": 4.8181818181818186e-05,
"loss": 2.3057,
"step": 7500
},
{
"epoch": 0.12,
"learning_rate": 4.806060606060606e-05,
"loss": 2.2837,
"step": 8000
},
{
"epoch": 0.12,
"learning_rate": 4.793939393939394e-05,
"loss": 2.2605,
"step": 8500
},
{
"epoch": 0.13,
"learning_rate": 4.781818181818182e-05,
"loss": 2.219,
"step": 9000
},
{
"epoch": 0.14,
"learning_rate": 4.76969696969697e-05,
"loss": 2.2163,
"step": 9500
},
{
"epoch": 0.15,
"learning_rate": 4.7575757575757576e-05,
"loss": 2.2109,
"step": 10000
},
{
"epoch": 0.15,
"learning_rate": 4.745454545454546e-05,
"loss": 2.1885,
"step": 10500
},
{
"epoch": 0.16,
"learning_rate": 4.7333333333333336e-05,
"loss": 2.1425,
"step": 11000
},
{
"epoch": 0.17,
"learning_rate": 4.721212121212121e-05,
"loss": 2.144,
"step": 11500
},
{
"epoch": 0.17,
"learning_rate": 4.709090909090909e-05,
"loss": 2.1339,
"step": 12000
},
{
"epoch": 0.18,
"learning_rate": 4.696969696969697e-05,
"loss": 2.1126,
"step": 12500
},
{
"epoch": 0.19,
"learning_rate": 4.684848484848485e-05,
"loss": 2.1071,
"step": 13000
},
{
"epoch": 0.2,
"learning_rate": 4.672727272727273e-05,
"loss": 2.1063,
"step": 13500
},
{
"epoch": 0.2,
"learning_rate": 4.660606060606061e-05,
"loss": 2.0812,
"step": 14000
},
{
"epoch": 0.21,
"learning_rate": 4.6484848484848487e-05,
"loss": 2.0689,
"step": 14500
},
{
"epoch": 0.22,
"learning_rate": 4.636363636363636e-05,
"loss": 2.0429,
"step": 15000
},
{
"epoch": 0.23,
"learning_rate": 4.624242424242424e-05,
"loss": 2.065,
"step": 15500
},
{
"epoch": 0.23,
"learning_rate": 4.612121212121212e-05,
"loss": 2.04,
"step": 16000
},
{
"epoch": 0.24,
"learning_rate": 4.600000000000001e-05,
"loss": 2.0448,
"step": 16500
},
{
"epoch": 0.25,
"learning_rate": 4.5878787878787883e-05,
"loss": 2.0126,
"step": 17000
},
{
"epoch": 0.25,
"learning_rate": 4.575757575757576e-05,
"loss": 2.0247,
"step": 17500
},
{
"epoch": 0.26,
"learning_rate": 4.563636363636364e-05,
"loss": 1.9993,
"step": 18000
},
{
"epoch": 0.27,
"learning_rate": 4.5515151515151513e-05,
"loss": 1.9819,
"step": 18500
},
{
"epoch": 0.28,
"learning_rate": 4.53939393939394e-05,
"loss": 1.9814,
"step": 19000
},
{
"epoch": 0.28,
"learning_rate": 4.5272727272727274e-05,
"loss": 1.9708,
"step": 19500
},
{
"epoch": 0.29,
"learning_rate": 4.515151515151516e-05,
"loss": 1.978,
"step": 20000
},
{
"epoch": 0.3,
"learning_rate": 4.5030303030303034e-05,
"loss": 1.9544,
"step": 20500
},
{
"epoch": 0.31,
"learning_rate": 4.490909090909091e-05,
"loss": 1.9555,
"step": 21000
},
{
"epoch": 0.31,
"learning_rate": 4.478787878787879e-05,
"loss": 1.9297,
"step": 21500
},
{
"epoch": 0.32,
"learning_rate": 4.466666666666667e-05,
"loss": 1.9572,
"step": 22000
},
{
"epoch": 0.33,
"learning_rate": 4.454545454545455e-05,
"loss": 1.9244,
"step": 22500
},
{
"epoch": 0.33,
"learning_rate": 4.4424242424242424e-05,
"loss": 1.9098,
"step": 23000
},
{
"epoch": 0.34,
"learning_rate": 4.430303030303031e-05,
"loss": 1.9108,
"step": 23500
},
{
"epoch": 0.35,
"learning_rate": 4.4181818181818184e-05,
"loss": 1.9152,
"step": 24000
},
{
"epoch": 0.36,
"learning_rate": 4.406060606060606e-05,
"loss": 1.8981,
"step": 24500
},
{
"epoch": 0.36,
"learning_rate": 4.3939393939393944e-05,
"loss": 1.889,
"step": 25000
},
{
"epoch": 0.37,
"learning_rate": 4.381818181818182e-05,
"loss": 1.8922,
"step": 25500
},
{
"epoch": 0.38,
"learning_rate": 4.36969696969697e-05,
"loss": 1.8827,
"step": 26000
},
{
"epoch": 0.39,
"learning_rate": 4.3575757575757574e-05,
"loss": 1.8801,
"step": 26500
},
{
"epoch": 0.39,
"learning_rate": 4.345454545454546e-05,
"loss": 1.8639,
"step": 27000
},
{
"epoch": 0.4,
"learning_rate": 4.3333333333333334e-05,
"loss": 1.8618,
"step": 27500
},
{
"epoch": 0.41,
"learning_rate": 4.321212121212122e-05,
"loss": 1.8786,
"step": 28000
},
{
"epoch": 0.41,
"learning_rate": 4.3090909090909094e-05,
"loss": 1.8485,
"step": 28500
},
{
"epoch": 0.42,
"learning_rate": 4.296969696969697e-05,
"loss": 1.8448,
"step": 29000
},
{
"epoch": 0.43,
"learning_rate": 4.284848484848485e-05,
"loss": 1.83,
"step": 29500
},
{
"epoch": 0.44,
"learning_rate": 4.2727272727272724e-05,
"loss": 1.8522,
"step": 30000
},
{
"epoch": 0.44,
"learning_rate": 4.260606060606061e-05,
"loss": 1.8352,
"step": 30500
},
{
"epoch": 0.45,
"learning_rate": 4.248484848484849e-05,
"loss": 1.8141,
"step": 31000
},
{
"epoch": 0.46,
"learning_rate": 4.236363636363637e-05,
"loss": 1.817,
"step": 31500
},
{
"epoch": 0.47,
"learning_rate": 4.2242424242424244e-05,
"loss": 1.8041,
"step": 32000
},
{
"epoch": 0.47,
"learning_rate": 4.212121212121212e-05,
"loss": 1.8027,
"step": 32500
},
{
"epoch": 0.48,
"learning_rate": 4.2e-05,
"loss": 1.8092,
"step": 33000
},
{
"epoch": 0.49,
"learning_rate": 4.187878787878788e-05,
"loss": 1.7975,
"step": 33500
},
{
"epoch": 0.49,
"learning_rate": 4.1757575757575765e-05,
"loss": 1.7954,
"step": 34000
},
{
"epoch": 0.5,
"learning_rate": 4.163636363636364e-05,
"loss": 1.7862,
"step": 34500
},
{
"epoch": 0.51,
"learning_rate": 4.151515151515152e-05,
"loss": 1.7732,
"step": 35000
},
{
"epoch": 0.52,
"learning_rate": 4.1393939393939395e-05,
"loss": 1.7512,
"step": 35500
},
{
"epoch": 0.52,
"learning_rate": 4.127272727272727e-05,
"loss": 1.7578,
"step": 36000
},
{
"epoch": 0.53,
"learning_rate": 4.1151515151515155e-05,
"loss": 1.7558,
"step": 36500
},
{
"epoch": 0.54,
"learning_rate": 4.103030303030303e-05,
"loss": 1.7377,
"step": 37000
},
{
"epoch": 0.55,
"learning_rate": 4.0909090909090915e-05,
"loss": 1.752,
"step": 37500
},
{
"epoch": 0.55,
"learning_rate": 4.078787878787879e-05,
"loss": 1.7293,
"step": 38000
},
{
"epoch": 0.56,
"learning_rate": 4.066666666666667e-05,
"loss": 1.7322,
"step": 38500
},
{
"epoch": 0.57,
"learning_rate": 4.0545454545454545e-05,
"loss": 1.7293,
"step": 39000
},
{
"epoch": 0.57,
"learning_rate": 4.042424242424243e-05,
"loss": 1.7348,
"step": 39500
},
{
"epoch": 0.58,
"learning_rate": 4.0303030303030305e-05,
"loss": 1.7133,
"step": 40000
},
{
"epoch": 0.59,
"learning_rate": 4.018181818181818e-05,
"loss": 1.7224,
"step": 40500
},
{
"epoch": 0.6,
"learning_rate": 4.0060606060606065e-05,
"loss": 1.7395,
"step": 41000
},
{
"epoch": 0.6,
"learning_rate": 3.993939393939394e-05,
"loss": 1.7224,
"step": 41500
},
{
"epoch": 0.61,
"learning_rate": 3.981818181818182e-05,
"loss": 1.7224,
"step": 42000
},
{
"epoch": 0.62,
"learning_rate": 3.96969696969697e-05,
"loss": 1.6872,
"step": 42500
},
{
"epoch": 0.63,
"learning_rate": 3.957575757575758e-05,
"loss": 1.7039,
"step": 43000
},
{
"epoch": 0.63,
"learning_rate": 3.9454545454545455e-05,
"loss": 1.7014,
"step": 43500
},
{
"epoch": 0.64,
"learning_rate": 3.933333333333333e-05,
"loss": 1.6762,
"step": 44000
},
{
"epoch": 0.65,
"learning_rate": 3.9212121212121215e-05,
"loss": 1.688,
"step": 44500
},
{
"epoch": 0.65,
"learning_rate": 3.909090909090909e-05,
"loss": 1.6812,
"step": 45000
},
{
"epoch": 0.66,
"learning_rate": 3.8969696969696975e-05,
"loss": 1.6776,
"step": 45500
},
{
"epoch": 0.67,
"learning_rate": 3.884848484848485e-05,
"loss": 1.6639,
"step": 46000
},
{
"epoch": 0.68,
"learning_rate": 3.872727272727273e-05,
"loss": 1.6613,
"step": 46500
},
{
"epoch": 0.68,
"learning_rate": 3.8606060606060605e-05,
"loss": 1.6907,
"step": 47000
},
{
"epoch": 0.69,
"learning_rate": 3.848484848484848e-05,
"loss": 1.6505,
"step": 47500
},
{
"epoch": 0.7,
"learning_rate": 3.8363636363636365e-05,
"loss": 1.6567,
"step": 48000
},
{
"epoch": 0.71,
"learning_rate": 3.824242424242425e-05,
"loss": 1.642,
"step": 48500
},
{
"epoch": 0.71,
"learning_rate": 3.8121212121212125e-05,
"loss": 1.6529,
"step": 49000
},
{
"epoch": 0.72,
"learning_rate": 3.8e-05,
"loss": 1.6652,
"step": 49500
},
{
"epoch": 0.73,
"learning_rate": 3.787878787878788e-05,
"loss": 1.6205,
"step": 50000
},
{
"epoch": 0.73,
"learning_rate": 3.7757575757575755e-05,
"loss": 1.6396,
"step": 50500
},
{
"epoch": 0.74,
"learning_rate": 3.763636363636364e-05,
"loss": 1.6264,
"step": 51000
},
{
"epoch": 0.75,
"learning_rate": 3.7515151515151516e-05,
"loss": 1.6381,
"step": 51500
},
{
"epoch": 0.76,
"learning_rate": 3.73939393939394e-05,
"loss": 1.6384,
"step": 52000
},
{
"epoch": 0.76,
"learning_rate": 3.7272727272727276e-05,
"loss": 1.6215,
"step": 52500
},
{
"epoch": 0.77,
"learning_rate": 3.715151515151515e-05,
"loss": 1.6265,
"step": 53000
},
{
"epoch": 0.78,
"learning_rate": 3.703030303030303e-05,
"loss": 1.6105,
"step": 53500
},
{
"epoch": 0.79,
"learning_rate": 3.690909090909091e-05,
"loss": 1.621,
"step": 54000
},
{
"epoch": 0.79,
"learning_rate": 3.678787878787879e-05,
"loss": 1.5984,
"step": 54500
},
{
"epoch": 0.8,
"learning_rate": 3.6666666666666666e-05,
"loss": 1.5843,
"step": 55000
},
{
"epoch": 0.81,
"learning_rate": 3.654545454545455e-05,
"loss": 1.5924,
"step": 55500
},
{
"epoch": 0.81,
"learning_rate": 3.6424242424242426e-05,
"loss": 1.6088,
"step": 56000
},
{
"epoch": 0.82,
"learning_rate": 3.63030303030303e-05,
"loss": 1.608,
"step": 56500
},
{
"epoch": 0.83,
"learning_rate": 3.6181818181818186e-05,
"loss": 1.5497,
"step": 57000
},
{
"epoch": 0.84,
"learning_rate": 3.606060606060606e-05,
"loss": 1.5803,
"step": 57500
},
{
"epoch": 0.84,
"learning_rate": 3.593939393939394e-05,
"loss": 1.5805,
"step": 58000
},
{
"epoch": 0.85,
"learning_rate": 3.5818181818181816e-05,
"loss": 1.5827,
"step": 58500
},
{
"epoch": 0.86,
"learning_rate": 3.56969696969697e-05,
"loss": 1.5819,
"step": 59000
},
{
"epoch": 0.87,
"learning_rate": 3.5575757575757576e-05,
"loss": 1.5652,
"step": 59500
},
{
"epoch": 0.87,
"learning_rate": 3.545454545454546e-05,
"loss": 1.5402,
"step": 60000
},
{
"epoch": 0.88,
"learning_rate": 3.5333333333333336e-05,
"loss": 1.5669,
"step": 60500
},
{
"epoch": 0.89,
"learning_rate": 3.521212121212121e-05,
"loss": 1.5428,
"step": 61000
},
{
"epoch": 0.89,
"learning_rate": 3.509090909090909e-05,
"loss": 1.5719,
"step": 61500
},
{
"epoch": 0.9,
"learning_rate": 3.4969696969696966e-05,
"loss": 1.5529,
"step": 62000
},
{
"epoch": 0.91,
"learning_rate": 3.484848484848485e-05,
"loss": 1.5315,
"step": 62500
},
{
"epoch": 0.92,
"learning_rate": 3.472727272727273e-05,
"loss": 1.5409,
"step": 63000
},
{
"epoch": 0.92,
"learning_rate": 3.460606060606061e-05,
"loss": 1.5517,
"step": 63500
},
{
"epoch": 0.93,
"learning_rate": 3.4484848484848486e-05,
"loss": 1.5464,
"step": 64000
},
{
"epoch": 0.94,
"learning_rate": 3.436363636363636e-05,
"loss": 1.5478,
"step": 64500
},
{
"epoch": 0.95,
"learning_rate": 3.424242424242424e-05,
"loss": 1.5304,
"step": 65000
},
{
"epoch": 0.95,
"learning_rate": 3.412121212121212e-05,
"loss": 1.5287,
"step": 65500
},
{
"epoch": 0.96,
"learning_rate": 3.4000000000000007e-05,
"loss": 1.5111,
"step": 66000
},
{
"epoch": 0.97,
"learning_rate": 3.387878787878788e-05,
"loss": 1.5287,
"step": 66500
},
{
"epoch": 0.97,
"learning_rate": 3.375757575757576e-05,
"loss": 1.5192,
"step": 67000
},
{
"epoch": 0.98,
"learning_rate": 3.3636363636363636e-05,
"loss": 1.5208,
"step": 67500
},
{
"epoch": 0.99,
"learning_rate": 3.351515151515151e-05,
"loss": 1.5139,
"step": 68000
},
{
"epoch": 1.0,
"learning_rate": 3.3393939393939397e-05,
"loss": 1.5067,
"step": 68500
},
{
"epoch": 1.0,
"eval_loss": 1.6837466955184937,
"eval_runtime": 1235.7847,
"eval_samples_per_second": 97.104,
"eval_steps_per_second": 3.035,
"step": 68750
},
{
"epoch": 1.0,
"learning_rate": 3.327272727272727e-05,
"loss": 1.4235,
"step": 69000
},
{
"epoch": 1.01,
"learning_rate": 3.315151515151516e-05,
"loss": 1.3688,
"step": 69500
},
{
"epoch": 1.02,
"learning_rate": 3.303030303030303e-05,
"loss": 1.3608,
"step": 70000
},
{
"epoch": 1.03,
"learning_rate": 3.290909090909091e-05,
"loss": 1.3435,
"step": 70500
},
{
"epoch": 1.03,
"learning_rate": 3.278787878787879e-05,
"loss": 1.3613,
"step": 71000
},
{
"epoch": 1.04,
"learning_rate": 3.266666666666667e-05,
"loss": 1.3527,
"step": 71500
},
{
"epoch": 1.05,
"learning_rate": 3.254545454545455e-05,
"loss": 1.3412,
"step": 72000
},
{
"epoch": 1.05,
"learning_rate": 3.2424242424242423e-05,
"loss": 1.3241,
"step": 72500
},
{
"epoch": 1.06,
"learning_rate": 3.230303030303031e-05,
"loss": 1.357,
"step": 73000
},
{
"epoch": 1.07,
"learning_rate": 3.2181818181818184e-05,
"loss": 1.3515,
"step": 73500
},
{
"epoch": 1.08,
"learning_rate": 3.206060606060606e-05,
"loss": 1.349,
"step": 74000
},
{
"epoch": 1.08,
"learning_rate": 3.1939393939393944e-05,
"loss": 1.3457,
"step": 74500
},
{
"epoch": 1.09,
"learning_rate": 3.181818181818182e-05,
"loss": 1.3598,
"step": 75000
},
{
"epoch": 1.1,
"learning_rate": 3.16969696969697e-05,
"loss": 1.3438,
"step": 75500
},
{
"epoch": 1.11,
"learning_rate": 3.1575757575757574e-05,
"loss": 1.333,
"step": 76000
},
{
"epoch": 1.11,
"learning_rate": 3.145454545454546e-05,
"loss": 1.3559,
"step": 76500
},
{
"epoch": 1.12,
"learning_rate": 3.1333333333333334e-05,
"loss": 1.3417,
"step": 77000
},
{
"epoch": 1.13,
"learning_rate": 3.121212121212122e-05,
"loss": 1.3425,
"step": 77500
},
{
"epoch": 1.13,
"learning_rate": 3.1090909090909094e-05,
"loss": 1.3549,
"step": 78000
},
{
"epoch": 1.14,
"learning_rate": 3.096969696969697e-05,
"loss": 1.3508,
"step": 78500
},
{
"epoch": 1.15,
"learning_rate": 3.084848484848485e-05,
"loss": 1.3481,
"step": 79000
},
{
"epoch": 1.16,
"learning_rate": 3.0727272727272724e-05,
"loss": 1.3256,
"step": 79500
},
{
"epoch": 1.16,
"learning_rate": 3.060606060606061e-05,
"loss": 1.3316,
"step": 80000
},
{
"epoch": 1.17,
"learning_rate": 3.0484848484848487e-05,
"loss": 1.3434,
"step": 80500
},
{
"epoch": 1.18,
"learning_rate": 3.0363636363636367e-05,
"loss": 1.3265,
"step": 81000
},
{
"epoch": 1.19,
"learning_rate": 3.0242424242424244e-05,
"loss": 1.3174,
"step": 81500
},
{
"epoch": 1.19,
"learning_rate": 3.012121212121212e-05,
"loss": 1.3155,
"step": 82000
},
{
"epoch": 1.2,
"learning_rate": 3e-05,
"loss": 1.3222,
"step": 82500
},
{
"epoch": 1.21,
"learning_rate": 2.9878787878787877e-05,
"loss": 1.3259,
"step": 83000
},
{
"epoch": 1.21,
"learning_rate": 2.9757575757575757e-05,
"loss": 1.3184,
"step": 83500
},
{
"epoch": 1.22,
"learning_rate": 2.963636363636364e-05,
"loss": 1.3262,
"step": 84000
},
{
"epoch": 1.23,
"learning_rate": 2.9515151515151518e-05,
"loss": 1.3059,
"step": 84500
},
{
"epoch": 1.24,
"learning_rate": 2.9393939393939394e-05,
"loss": 1.3129,
"step": 85000
},
{
"epoch": 1.24,
"learning_rate": 2.9272727272727274e-05,
"loss": 1.3068,
"step": 85500
},
{
"epoch": 1.25,
"learning_rate": 2.915151515151515e-05,
"loss": 1.3208,
"step": 86000
},
{
"epoch": 1.26,
"learning_rate": 2.903030303030303e-05,
"loss": 1.298,
"step": 86500
},
{
"epoch": 1.27,
"learning_rate": 2.8909090909090908e-05,
"loss": 1.3132,
"step": 87000
},
{
"epoch": 1.27,
"learning_rate": 2.878787878787879e-05,
"loss": 1.3159,
"step": 87500
},
{
"epoch": 1.28,
"learning_rate": 2.8666666666666668e-05,
"loss": 1.3063,
"step": 88000
},
{
"epoch": 1.29,
"learning_rate": 2.8545454545454548e-05,
"loss": 1.3065,
"step": 88500
},
{
"epoch": 1.29,
"learning_rate": 2.8424242424242424e-05,
"loss": 1.2907,
"step": 89000
},
{
"epoch": 1.3,
"learning_rate": 2.8303030303030305e-05,
"loss": 1.3093,
"step": 89500
},
{
"epoch": 1.31,
"learning_rate": 2.818181818181818e-05,
"loss": 1.3054,
"step": 90000
},
{
"epoch": 1.32,
"learning_rate": 2.8060606060606058e-05,
"loss": 1.3046,
"step": 90500
},
{
"epoch": 1.32,
"learning_rate": 2.7939393939393945e-05,
"loss": 1.3251,
"step": 91000
},
{
"epoch": 1.33,
"learning_rate": 2.781818181818182e-05,
"loss": 1.2915,
"step": 91500
},
{
"epoch": 1.34,
"learning_rate": 2.7696969696969698e-05,
"loss": 1.3113,
"step": 92000
},
{
"epoch": 1.35,
"learning_rate": 2.7575757575757578e-05,
"loss": 1.2773,
"step": 92500
},
{
"epoch": 1.35,
"learning_rate": 2.7454545454545455e-05,
"loss": 1.3076,
"step": 93000
},
{
"epoch": 1.36,
"learning_rate": 2.733333333333333e-05,
"loss": 1.2772,
"step": 93500
},
{
"epoch": 1.37,
"learning_rate": 2.721212121212121e-05,
"loss": 1.275,
"step": 94000
},
{
"epoch": 1.37,
"learning_rate": 2.7090909090909095e-05,
"loss": 1.2929,
"step": 94500
},
{
"epoch": 1.38,
"learning_rate": 2.696969696969697e-05,
"loss": 1.2849,
"step": 95000
},
{
"epoch": 1.39,
"learning_rate": 2.684848484848485e-05,
"loss": 1.2681,
"step": 95500
},
{
"epoch": 1.4,
"learning_rate": 2.6727272727272728e-05,
"loss": 1.303,
"step": 96000
},
{
"epoch": 1.4,
"learning_rate": 2.6606060606060605e-05,
"loss": 1.2876,
"step": 96500
},
{
"epoch": 1.41,
"learning_rate": 2.6484848484848485e-05,
"loss": 1.2657,
"step": 97000
},
{
"epoch": 1.42,
"learning_rate": 2.636363636363636e-05,
"loss": 1.28,
"step": 97500
},
{
"epoch": 1.43,
"learning_rate": 2.6242424242424245e-05,
"loss": 1.2482,
"step": 98000
},
{
"epoch": 1.43,
"learning_rate": 2.6121212121212125e-05,
"loss": 1.2574,
"step": 98500
},
{
"epoch": 1.44,
"learning_rate": 2.6000000000000002e-05,
"loss": 1.2557,
"step": 99000
},
{
"epoch": 1.45,
"learning_rate": 2.587878787878788e-05,
"loss": 1.2827,
"step": 99500
},
{
"epoch": 1.45,
"learning_rate": 2.575757575757576e-05,
"loss": 1.2572,
"step": 100000
},
{
"epoch": 1.46,
"learning_rate": 2.5636363636363635e-05,
"loss": 1.2571,
"step": 100500
},
{
"epoch": 1.47,
"learning_rate": 2.5515151515151515e-05,
"loss": 1.2707,
"step": 101000
},
{
"epoch": 1.48,
"learning_rate": 2.53939393939394e-05,
"loss": 1.2468,
"step": 101500
},
{
"epoch": 1.48,
"learning_rate": 2.5272727272727275e-05,
"loss": 1.2447,
"step": 102000
},
{
"epoch": 1.49,
"learning_rate": 2.5151515151515155e-05,
"loss": 1.2617,
"step": 102500
},
{
"epoch": 1.5,
"learning_rate": 2.5030303030303032e-05,
"loss": 1.2277,
"step": 103000
},
{
"epoch": 1.51,
"learning_rate": 2.490909090909091e-05,
"loss": 1.2725,
"step": 103500
},
{
"epoch": 1.51,
"learning_rate": 2.478787878787879e-05,
"loss": 1.2358,
"step": 104000
},
{
"epoch": 1.52,
"learning_rate": 2.466666666666667e-05,
"loss": 1.241,
"step": 104500
},
{
"epoch": 1.53,
"learning_rate": 2.4545454545454545e-05,
"loss": 1.249,
"step": 105000
},
{
"epoch": 1.53,
"learning_rate": 2.4424242424242426e-05,
"loss": 1.2212,
"step": 105500
},
{
"epoch": 1.54,
"learning_rate": 2.4303030303030306e-05,
"loss": 1.2323,
"step": 106000
},
{
"epoch": 1.55,
"learning_rate": 2.4181818181818182e-05,
"loss": 1.228,
"step": 106500
},
{
"epoch": 1.56,
"learning_rate": 2.4060606060606062e-05,
"loss": 1.2124,
"step": 107000
},
{
"epoch": 1.56,
"learning_rate": 2.393939393939394e-05,
"loss": 1.2512,
"step": 107500
},
{
"epoch": 1.57,
"learning_rate": 2.381818181818182e-05,
"loss": 1.2188,
"step": 108000
},
{
"epoch": 1.58,
"learning_rate": 2.36969696969697e-05,
"loss": 1.217,
"step": 108500
},
{
"epoch": 1.59,
"learning_rate": 2.3575757575757576e-05,
"loss": 1.2696,
"step": 109000
},
{
"epoch": 1.59,
"learning_rate": 2.3454545454545456e-05,
"loss": 1.223,
"step": 109500
},
{
"epoch": 1.6,
"learning_rate": 2.3333333333333336e-05,
"loss": 1.2262,
"step": 110000
},
{
"epoch": 1.61,
"learning_rate": 2.3212121212121212e-05,
"loss": 1.2129,
"step": 110500
},
{
"epoch": 1.61,
"learning_rate": 2.309090909090909e-05,
"loss": 1.2198,
"step": 111000
},
{
"epoch": 1.62,
"learning_rate": 2.2969696969696973e-05,
"loss": 1.2001,
"step": 111500
},
{
"epoch": 1.63,
"learning_rate": 2.284848484848485e-05,
"loss": 1.2129,
"step": 112000
},
{
"epoch": 1.64,
"learning_rate": 2.272727272727273e-05,
"loss": 1.2044,
"step": 112500
},
{
"epoch": 1.64,
"learning_rate": 2.260606060606061e-05,
"loss": 1.2219,
"step": 113000
},
{
"epoch": 1.65,
"learning_rate": 2.2484848484848486e-05,
"loss": 1.2214,
"step": 113500
},
{
"epoch": 1.66,
"learning_rate": 2.2363636363636366e-05,
"loss": 1.1948,
"step": 114000
},
{
"epoch": 1.67,
"learning_rate": 2.2242424242424243e-05,
"loss": 1.2041,
"step": 114500
},
{
"epoch": 1.67,
"learning_rate": 2.2121212121212123e-05,
"loss": 1.2114,
"step": 115000
},
{
"epoch": 1.68,
"learning_rate": 2.2000000000000003e-05,
"loss": 1.2051,
"step": 115500
},
{
"epoch": 1.69,
"learning_rate": 2.187878787878788e-05,
"loss": 1.2059,
"step": 116000
},
{
"epoch": 1.69,
"learning_rate": 2.175757575757576e-05,
"loss": 1.2335,
"step": 116500
},
{
"epoch": 1.7,
"learning_rate": 2.163636363636364e-05,
"loss": 1.2088,
"step": 117000
},
{
"epoch": 1.71,
"learning_rate": 2.1515151515151516e-05,
"loss": 1.2058,
"step": 117500
},
{
"epoch": 1.72,
"learning_rate": 2.1393939393939393e-05,
"loss": 1.1982,
"step": 118000
},
{
"epoch": 1.72,
"learning_rate": 2.1272727272727276e-05,
"loss": 1.1952,
"step": 118500
},
{
"epoch": 1.73,
"learning_rate": 2.1151515151515153e-05,
"loss": 1.1729,
"step": 119000
},
{
"epoch": 1.74,
"learning_rate": 2.103030303030303e-05,
"loss": 1.1796,
"step": 119500
},
{
"epoch": 1.75,
"learning_rate": 2.090909090909091e-05,
"loss": 1.1817,
"step": 120000
},
{
"epoch": 1.75,
"learning_rate": 2.078787878787879e-05,
"loss": 1.1833,
"step": 120500
},
{
"epoch": 1.76,
"learning_rate": 2.0666666666666666e-05,
"loss": 1.182,
"step": 121000
},
{
"epoch": 1.77,
"learning_rate": 2.0545454545454546e-05,
"loss": 1.1816,
"step": 121500
},
{
"epoch": 1.77,
"learning_rate": 2.0424242424242427e-05,
"loss": 1.1666,
"step": 122000
},
{
"epoch": 1.78,
"learning_rate": 2.0303030303030303e-05,
"loss": 1.1642,
"step": 122500
},
{
"epoch": 1.79,
"learning_rate": 2.0181818181818183e-05,
"loss": 1.1658,
"step": 123000
},
{
"epoch": 1.8,
"learning_rate": 2.006060606060606e-05,
"loss": 1.1708,
"step": 123500
},
{
"epoch": 1.8,
"learning_rate": 1.993939393939394e-05,
"loss": 1.1742,
"step": 124000
},
{
"epoch": 1.81,
"learning_rate": 1.981818181818182e-05,
"loss": 1.1614,
"step": 124500
},
{
"epoch": 1.82,
"learning_rate": 1.9696969696969697e-05,
"loss": 1.17,
"step": 125000
},
{
"epoch": 1.83,
"learning_rate": 1.9575757575757577e-05,
"loss": 1.1679,
"step": 125500
},
{
"epoch": 1.83,
"learning_rate": 1.9454545454545457e-05,
"loss": 1.1545,
"step": 126000
},
{
"epoch": 1.84,
"learning_rate": 1.9333333333333333e-05,
"loss": 1.1688,
"step": 126500
},
{
"epoch": 1.85,
"learning_rate": 1.9212121212121213e-05,
"loss": 1.1615,
"step": 127000
},
{
"epoch": 1.85,
"learning_rate": 1.9090909090909094e-05,
"loss": 1.1366,
"step": 127500
},
{
"epoch": 1.86,
"learning_rate": 1.896969696969697e-05,
"loss": 1.1694,
"step": 128000
},
{
"epoch": 1.87,
"learning_rate": 1.884848484848485e-05,
"loss": 1.1326,
"step": 128500
},
{
"epoch": 1.88,
"learning_rate": 1.872727272727273e-05,
"loss": 1.1274,
"step": 129000
},
{
"epoch": 1.88,
"learning_rate": 1.8606060606060607e-05,
"loss": 1.1415,
"step": 129500
},
{
"epoch": 1.89,
"learning_rate": 1.8484848484848487e-05,
"loss": 1.1488,
"step": 130000
},
{
"epoch": 1.9,
"learning_rate": 1.8363636363636364e-05,
"loss": 1.1248,
"step": 130500
},
{
"epoch": 1.91,
"learning_rate": 1.8242424242424244e-05,
"loss": 1.1589,
"step": 131000
},
{
"epoch": 1.91,
"learning_rate": 1.8121212121212124e-05,
"loss": 1.133,
"step": 131500
},
{
"epoch": 1.92,
"learning_rate": 1.8e-05,
"loss": 1.1291,
"step": 132000
},
{
"epoch": 1.93,
"learning_rate": 1.787878787878788e-05,
"loss": 1.1552,
"step": 132500
},
{
"epoch": 1.93,
"learning_rate": 1.775757575757576e-05,
"loss": 1.1193,
"step": 133000
},
{
"epoch": 1.94,
"learning_rate": 1.7636363636363637e-05,
"loss": 1.1384,
"step": 133500
},
{
"epoch": 1.95,
"learning_rate": 1.7515151515151514e-05,
"loss": 1.1292,
"step": 134000
},
{
"epoch": 1.96,
"learning_rate": 1.7393939393939397e-05,
"loss": 1.1318,
"step": 134500
},
{
"epoch": 1.96,
"learning_rate": 1.7272727272727274e-05,
"loss": 1.1061,
"step": 135000
},
{
"epoch": 1.97,
"learning_rate": 1.715151515151515e-05,
"loss": 1.1182,
"step": 135500
},
{
"epoch": 1.98,
"learning_rate": 1.703030303030303e-05,
"loss": 1.1277,
"step": 136000
},
{
"epoch": 1.99,
"learning_rate": 1.690909090909091e-05,
"loss": 1.1243,
"step": 136500
},
{
"epoch": 1.99,
"learning_rate": 1.6787878787878787e-05,
"loss": 1.1209,
"step": 137000
},
{
"epoch": 2.0,
"learning_rate": 1.6666666666666667e-05,
"loss": 1.1444,
"step": 137500
},
{
"epoch": 2.0,
"eval_loss": 1.3132866621017456,
"eval_runtime": 1353.544,
"eval_samples_per_second": 88.656,
"eval_steps_per_second": 2.771,
"step": 137500
},
{
"epoch": 2.01,
"learning_rate": 1.6545454545454548e-05,
"loss": 0.8966,
"step": 138000
},
{
"epoch": 2.01,
"learning_rate": 1.6424242424242424e-05,
"loss": 0.9101,
"step": 138500
},
{
"epoch": 2.02,
"learning_rate": 1.6303030303030304e-05,
"loss": 0.911,
"step": 139000
},
{
"epoch": 2.03,
"learning_rate": 1.618181818181818e-05,
"loss": 0.896,
"step": 139500
},
{
"epoch": 2.04,
"learning_rate": 1.606060606060606e-05,
"loss": 0.9047,
"step": 140000
},
{
"epoch": 2.04,
"learning_rate": 1.593939393939394e-05,
"loss": 0.891,
"step": 140500
},
{
"epoch": 2.05,
"learning_rate": 1.5818181818181818e-05,
"loss": 0.8898,
"step": 141000
},
{
"epoch": 2.06,
"learning_rate": 1.5696969696969698e-05,
"loss": 0.8982,
"step": 141500
},
{
"epoch": 2.07,
"learning_rate": 1.5575757575757578e-05,
"loss": 0.8965,
"step": 142000
},
{
"epoch": 2.07,
"learning_rate": 1.5454545454545454e-05,
"loss": 0.8793,
"step": 142500
},
{
"epoch": 2.08,
"learning_rate": 1.5333333333333334e-05,
"loss": 0.8738,
"step": 143000
},
{
"epoch": 2.09,
"learning_rate": 1.5212121212121213e-05,
"loss": 0.8893,
"step": 143500
},
{
"epoch": 2.09,
"learning_rate": 1.5090909090909091e-05,
"loss": 0.8802,
"step": 144000
},
{
"epoch": 2.1,
"learning_rate": 1.496969696969697e-05,
"loss": 0.8884,
"step": 144500
},
{
"epoch": 2.11,
"learning_rate": 1.484848484848485e-05,
"loss": 0.8822,
"step": 145000
},
{
"epoch": 2.12,
"learning_rate": 1.4727272727272728e-05,
"loss": 0.8881,
"step": 145500
},
{
"epoch": 2.12,
"learning_rate": 1.4606060606060606e-05,
"loss": 0.8865,
"step": 146000
},
{
"epoch": 2.13,
"learning_rate": 1.4484848484848485e-05,
"loss": 0.8804,
"step": 146500
},
{
"epoch": 2.14,
"learning_rate": 1.4363636363636365e-05,
"loss": 0.8792,
"step": 147000
},
{
"epoch": 2.15,
"learning_rate": 1.4242424242424243e-05,
"loss": 0.8752,
"step": 147500
},
{
"epoch": 2.15,
"learning_rate": 1.4121212121212121e-05,
"loss": 0.8634,
"step": 148000
},
{
"epoch": 2.16,
"learning_rate": 1.4000000000000001e-05,
"loss": 0.897,
"step": 148500
},
{
"epoch": 2.17,
"learning_rate": 1.387878787878788e-05,
"loss": 0.8875,
"step": 149000
},
{
"epoch": 2.17,
"learning_rate": 1.3757575757575758e-05,
"loss": 0.8764,
"step": 149500
},
{
"epoch": 2.18,
"learning_rate": 1.3636363636363637e-05,
"loss": 0.8813,
"step": 150000
},
{
"epoch": 2.19,
"learning_rate": 1.3515151515151517e-05,
"loss": 0.885,
"step": 150500
},
{
"epoch": 2.2,
"learning_rate": 1.3393939393939395e-05,
"loss": 0.8793,
"step": 151000
},
{
"epoch": 2.2,
"learning_rate": 1.3272727272727273e-05,
"loss": 0.8677,
"step": 151500
},
{
"epoch": 2.21,
"learning_rate": 1.315151515151515e-05,
"loss": 0.8783,
"step": 152000
},
{
"epoch": 2.22,
"learning_rate": 1.3030303030303032e-05,
"loss": 0.8702,
"step": 152500
},
{
"epoch": 2.23,
"learning_rate": 1.290909090909091e-05,
"loss": 0.8714,
"step": 153000
},
{
"epoch": 2.23,
"learning_rate": 1.2787878787878787e-05,
"loss": 0.8797,
"step": 153500
},
{
"epoch": 2.24,
"learning_rate": 1.2666666666666668e-05,
"loss": 0.8617,
"step": 154000
},
{
"epoch": 2.25,
"learning_rate": 1.2545454545454547e-05,
"loss": 0.8746,
"step": 154500
},
{
"epoch": 2.25,
"learning_rate": 1.2424242424242424e-05,
"loss": 0.868,
"step": 155000
},
{
"epoch": 2.26,
"learning_rate": 1.2303030303030304e-05,
"loss": 0.8676,
"step": 155500
},
{
"epoch": 2.27,
"learning_rate": 1.2181818181818182e-05,
"loss": 0.8562,
"step": 156000
},
{
"epoch": 2.28,
"learning_rate": 1.2060606060606062e-05,
"loss": 0.8467,
"step": 156500
},
{
"epoch": 2.28,
"learning_rate": 1.193939393939394e-05,
"loss": 0.8588,
"step": 157000
},
{
"epoch": 2.29,
"learning_rate": 1.1818181818181819e-05,
"loss": 0.8715,
"step": 157500
},
{
"epoch": 2.3,
"learning_rate": 1.1696969696969699e-05,
"loss": 0.8486,
"step": 158000
},
{
"epoch": 2.31,
"learning_rate": 1.1575757575757575e-05,
"loss": 0.8546,
"step": 158500
},
{
"epoch": 2.31,
"learning_rate": 1.1454545454545455e-05,
"loss": 0.8556,
"step": 159000
},
{
"epoch": 2.32,
"learning_rate": 1.1333333333333334e-05,
"loss": 0.8471,
"step": 159500
},
{
"epoch": 2.33,
"learning_rate": 1.1212121212121212e-05,
"loss": 0.8617,
"step": 160000
},
{
"epoch": 2.33,
"learning_rate": 1.1090909090909092e-05,
"loss": 0.8625,
"step": 160500
},
{
"epoch": 2.34,
"learning_rate": 1.096969696969697e-05,
"loss": 0.8325,
"step": 161000
},
{
"epoch": 2.35,
"learning_rate": 1.0848484848484849e-05,
"loss": 0.8361,
"step": 161500
},
{
"epoch": 2.36,
"learning_rate": 1.0727272727272727e-05,
"loss": 0.8415,
"step": 162000
},
{
"epoch": 2.36,
"learning_rate": 1.0606060606060607e-05,
"loss": 0.8384,
"step": 162500
},
{
"epoch": 2.37,
"learning_rate": 1.0484848484848486e-05,
"loss": 0.8377,
"step": 163000
},
{
"epoch": 2.38,
"learning_rate": 1.0363636363636364e-05,
"loss": 0.8614,
"step": 163500
},
{
"epoch": 2.39,
"learning_rate": 1.0242424242424242e-05,
"loss": 0.836,
"step": 164000
},
{
"epoch": 2.39,
"learning_rate": 1.0121212121212122e-05,
"loss": 0.8498,
"step": 164500
},
{
"epoch": 2.4,
"learning_rate": 1e-05,
"loss": 0.8287,
"step": 165000
},
{
"epoch": 2.41,
"learning_rate": 9.87878787878788e-06,
"loss": 0.8294,
"step": 165500
},
{
"epoch": 2.41,
"learning_rate": 9.75757575757576e-06,
"loss": 0.836,
"step": 166000
},
{
"epoch": 2.42,
"learning_rate": 9.636363636363636e-06,
"loss": 0.8461,
"step": 166500
},
{
"epoch": 2.43,
"learning_rate": 9.515151515151516e-06,
"loss": 0.8331,
"step": 167000
},
{
"epoch": 2.44,
"learning_rate": 9.393939393939394e-06,
"loss": 0.8283,
"step": 167500
},
{
"epoch": 2.44,
"learning_rate": 9.272727272727273e-06,
"loss": 0.831,
"step": 168000
},
{
"epoch": 2.45,
"learning_rate": 9.151515151515153e-06,
"loss": 0.8443,
"step": 168500
},
{
"epoch": 2.46,
"learning_rate": 9.030303030303031e-06,
"loss": 0.8291,
"step": 169000
},
{
"epoch": 2.47,
"learning_rate": 8.90909090909091e-06,
"loss": 0.83,
"step": 169500
},
{
"epoch": 2.47,
"learning_rate": 8.787878787878788e-06,
"loss": 0.814,
"step": 170000
},
{
"epoch": 2.48,
"learning_rate": 8.666666666666668e-06,
"loss": 0.8187,
"step": 170500
},
{
"epoch": 2.49,
"learning_rate": 8.545454545454546e-06,
"loss": 0.8217,
"step": 171000
},
{
"epoch": 2.49,
"learning_rate": 8.424242424242425e-06,
"loss": 0.8017,
"step": 171500
},
{
"epoch": 2.5,
"learning_rate": 8.303030303030303e-06,
"loss": 0.8359,
"step": 172000
},
{
"epoch": 2.51,
"learning_rate": 8.181818181818183e-06,
"loss": 0.8134,
"step": 172500
},
{
"epoch": 2.52,
"learning_rate": 8.060606060606061e-06,
"loss": 0.8187,
"step": 173000
},
{
"epoch": 2.52,
"learning_rate": 7.93939393939394e-06,
"loss": 0.8198,
"step": 173500
},
{
"epoch": 2.53,
"learning_rate": 7.81818181818182e-06,
"loss": 0.8052,
"step": 174000
},
{
"epoch": 2.54,
"learning_rate": 7.696969696969696e-06,
"loss": 0.7909,
"step": 174500
},
{
"epoch": 2.55,
"learning_rate": 7.5757575757575764e-06,
"loss": 0.8181,
"step": 175000
},
{
"epoch": 2.55,
"learning_rate": 7.454545454545454e-06,
"loss": 0.8195,
"step": 175500
},
{
"epoch": 2.56,
"learning_rate": 7.333333333333334e-06,
"loss": 0.8166,
"step": 176000
},
{
"epoch": 2.57,
"learning_rate": 7.212121212121213e-06,
"loss": 0.8097,
"step": 176500
},
{
"epoch": 2.57,
"learning_rate": 7.090909090909091e-06,
"loss": 0.8263,
"step": 177000
},
{
"epoch": 2.58,
"learning_rate": 6.969696969696971e-06,
"loss": 0.8015,
"step": 177500
},
{
"epoch": 2.59,
"learning_rate": 6.848484848484848e-06,
"loss": 0.825,
"step": 178000
},
{
"epoch": 2.6,
"learning_rate": 6.727272727272728e-06,
"loss": 0.7864,
"step": 178500
},
{
"epoch": 2.6,
"learning_rate": 6.606060606060606e-06,
"loss": 0.7898,
"step": 179000
},
{
"epoch": 2.61,
"learning_rate": 6.484848484848485e-06,
"loss": 0.7985,
"step": 179500
},
{
"epoch": 2.62,
"learning_rate": 6.363636363636363e-06,
"loss": 0.7992,
"step": 180000
},
{
"epoch": 2.63,
"learning_rate": 6.242424242424243e-06,
"loss": 0.7927,
"step": 180500
},
{
"epoch": 2.63,
"learning_rate": 6.121212121212121e-06,
"loss": 0.7837,
"step": 181000
},
{
"epoch": 2.64,
"learning_rate": 6e-06,
"loss": 0.7979,
"step": 181500
},
{
"epoch": 2.65,
"learning_rate": 5.8787878787878785e-06,
"loss": 0.7898,
"step": 182000
},
{
"epoch": 2.65,
"learning_rate": 5.7575757575757586e-06,
"loss": 0.7875,
"step": 182500
},
{
"epoch": 2.66,
"learning_rate": 5.636363636363637e-06,
"loss": 0.785,
"step": 183000
},
{
"epoch": 2.67,
"learning_rate": 5.515151515151515e-06,
"loss": 0.7715,
"step": 183500
},
{
"epoch": 2.68,
"learning_rate": 5.3939393939393945e-06,
"loss": 0.7893,
"step": 184000
},
{
"epoch": 2.68,
"learning_rate": 5.272727272727273e-06,
"loss": 0.7781,
"step": 184500
},
{
"epoch": 2.69,
"learning_rate": 5.151515151515152e-06,
"loss": 0.7806,
"step": 185000
},
{
"epoch": 2.7,
"learning_rate": 5.03030303030303e-06,
"loss": 0.7867,
"step": 185500
},
{
"epoch": 2.71,
"learning_rate": 4.909090909090909e-06,
"loss": 0.7796,
"step": 186000
},
{
"epoch": 2.71,
"learning_rate": 4.787878787878789e-06,
"loss": 0.7635,
"step": 186500
},
{
"epoch": 2.72,
"learning_rate": 4.666666666666667e-06,
"loss": 0.7691,
"step": 187000
},
{
"epoch": 2.73,
"learning_rate": 4.5454545454545455e-06,
"loss": 0.7687,
"step": 187500
},
{
"epoch": 2.73,
"learning_rate": 4.424242424242425e-06,
"loss": 0.7758,
"step": 188000
},
{
"epoch": 2.74,
"learning_rate": 4.303030303030303e-06,
"loss": 0.7522,
"step": 188500
},
{
"epoch": 2.75,
"learning_rate": 4.181818181818182e-06,
"loss": 0.7664,
"step": 189000
},
{
"epoch": 2.76,
"learning_rate": 4.060606060606061e-06,
"loss": 0.7498,
"step": 189500
},
{
"epoch": 2.76,
"learning_rate": 3.939393939393939e-06,
"loss": 0.7549,
"step": 190000
},
{
"epoch": 2.77,
"learning_rate": 3.818181818181819e-06,
"loss": 0.7517,
"step": 190500
},
{
"epoch": 2.78,
"learning_rate": 3.6969696969696974e-06,
"loss": 0.7448,
"step": 191000
},
{
"epoch": 2.79,
"learning_rate": 3.575757575757576e-06,
"loss": 0.7567,
"step": 191500
},
{
"epoch": 2.79,
"learning_rate": 3.4545454545454545e-06,
"loss": 0.7617,
"step": 192000
},
{
"epoch": 2.8,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.7708,
"step": 192500
},
{
"epoch": 2.81,
"learning_rate": 3.212121212121212e-06,
"loss": 0.7508,
"step": 193000
},
{
"epoch": 2.81,
"learning_rate": 3.0909090909090913e-06,
"loss": 0.7575,
"step": 193500
},
{
"epoch": 2.82,
"learning_rate": 2.96969696969697e-06,
"loss": 0.7589,
"step": 194000
},
{
"epoch": 2.83,
"learning_rate": 2.8484848484848484e-06,
"loss": 0.7579,
"step": 194500
},
{
"epoch": 2.84,
"learning_rate": 2.7272727272727272e-06,
"loss": 0.7557,
"step": 195000
},
{
"epoch": 2.84,
"learning_rate": 2.6060606060606064e-06,
"loss": 0.7506,
"step": 195500
},
{
"epoch": 2.85,
"learning_rate": 2.484848484848485e-06,
"loss": 0.771,
"step": 196000
},
{
"epoch": 2.86,
"learning_rate": 2.3636363636363636e-06,
"loss": 0.7393,
"step": 196500
},
{
"epoch": 2.87,
"learning_rate": 2.2424242424242423e-06,
"loss": 0.7489,
"step": 197000
},
{
"epoch": 2.87,
"learning_rate": 2.1212121212121216e-06,
"loss": 0.7467,
"step": 197500
},
{
"epoch": 2.88,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.7489,
"step": 198000
},
{
"epoch": 2.89,
"learning_rate": 1.878787878787879e-06,
"loss": 0.7469,
"step": 198500
},
{
"epoch": 2.89,
"learning_rate": 1.7575757575757575e-06,
"loss": 0.7421,
"step": 199000
},
{
"epoch": 2.9,
"learning_rate": 1.6363636363636367e-06,
"loss": 0.7514,
"step": 199500
},
{
"epoch": 2.91,
"learning_rate": 1.5151515151515152e-06,
"loss": 0.75,
"step": 200000
},
{
"epoch": 2.92,
"learning_rate": 1.393939393939394e-06,
"loss": 0.7481,
"step": 200500
},
{
"epoch": 2.92,
"learning_rate": 1.2727272727272728e-06,
"loss": 0.7357,
"step": 201000
},
{
"epoch": 2.93,
"learning_rate": 1.1515151515151516e-06,
"loss": 0.7467,
"step": 201500
},
{
"epoch": 2.94,
"learning_rate": 1.0303030303030304e-06,
"loss": 0.7441,
"step": 202000
},
{
"epoch": 2.95,
"learning_rate": 9.09090909090909e-07,
"loss": 0.7428,
"step": 202500
},
{
"epoch": 2.95,
"learning_rate": 7.878787878787879e-07,
"loss": 0.7405,
"step": 203000
},
{
"epoch": 2.96,
"learning_rate": 6.666666666666667e-07,
"loss": 0.7483,
"step": 203500
},
{
"epoch": 2.97,
"learning_rate": 5.454545454545455e-07,
"loss": 0.7336,
"step": 204000
},
{
"epoch": 2.97,
"learning_rate": 4.242424242424243e-07,
"loss": 0.7378,
"step": 204500
},
{
"epoch": 2.98,
"learning_rate": 3.0303030303030305e-07,
"loss": 0.744,
"step": 205000
},
{
"epoch": 2.99,
"learning_rate": 1.8181818181818183e-07,
"loss": 0.7353,
"step": 205500
},
{
"epoch": 3.0,
"learning_rate": 6.060606060606061e-08,
"loss": 0.7653,
"step": 206000
},
{
"epoch": 3.0,
"eval_loss": 1.1062601804733276,
"eval_runtime": 1264.7443,
"eval_samples_per_second": 94.881,
"eval_steps_per_second": 2.965,
"step": 206250
},
{
"epoch": 3.0,
"step": 206250,
"total_flos": 6.62657819292053e+17,
"train_loss": 1.3104234486712831,
"train_runtime": 131784.1208,
"train_samples_per_second": 50.082,
"train_steps_per_second": 1.565
}
],
"max_steps": 206250,
"num_train_epochs": 3,
"total_flos": 6.62657819292053e+17,
"trial_name": null,
"trial_params": null
}