gpt2-345M-2.2B-USPTO / trainer_state.json
allyyc
files
401db01
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 68000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.9632352941176476e-05,
"loss": 6.0455,
"step": 500
},
{
"epoch": 0.01,
"learning_rate": 4.9264705882352944e-05,
"loss": 5.0041,
"step": 1000
},
{
"epoch": 0.02,
"learning_rate": 4.889705882352941e-05,
"loss": 4.4652,
"step": 1500
},
{
"epoch": 0.03,
"learning_rate": 4.8529411764705885e-05,
"loss": 4.0969,
"step": 2000
},
{
"epoch": 0.04,
"learning_rate": 4.816176470588236e-05,
"loss": 3.8476,
"step": 2500
},
{
"epoch": 0.04,
"learning_rate": 4.7794117647058826e-05,
"loss": 3.7641,
"step": 3000
},
{
"epoch": 0.05,
"learning_rate": 4.742647058823529e-05,
"loss": 3.5194,
"step": 3500
},
{
"epoch": 0.06,
"learning_rate": 4.705882352941177e-05,
"loss": 3.4048,
"step": 4000
},
{
"epoch": 0.07,
"learning_rate": 4.669117647058824e-05,
"loss": 3.3055,
"step": 4500
},
{
"epoch": 0.07,
"learning_rate": 4.632352941176471e-05,
"loss": 3.2199,
"step": 5000
},
{
"epoch": 0.08,
"learning_rate": 4.5955882352941176e-05,
"loss": 3.171,
"step": 5500
},
{
"epoch": 0.09,
"learning_rate": 4.558823529411765e-05,
"loss": 3.1226,
"step": 6000
},
{
"epoch": 0.1,
"learning_rate": 4.522058823529412e-05,
"loss": 3.0694,
"step": 6500
},
{
"epoch": 0.1,
"learning_rate": 4.485294117647059e-05,
"loss": 3.0365,
"step": 7000
},
{
"epoch": 0.11,
"learning_rate": 4.448529411764706e-05,
"loss": 3.0121,
"step": 7500
},
{
"epoch": 0.12,
"learning_rate": 4.411764705882353e-05,
"loss": 2.9746,
"step": 8000
},
{
"epoch": 0.12,
"learning_rate": 4.375e-05,
"loss": 2.9517,
"step": 8500
},
{
"epoch": 0.13,
"learning_rate": 4.3382352941176474e-05,
"loss": 2.9222,
"step": 9000
},
{
"epoch": 0.14,
"learning_rate": 4.301470588235295e-05,
"loss": 3.0136,
"step": 9500
},
{
"epoch": 0.15,
"learning_rate": 4.2647058823529415e-05,
"loss": 2.9355,
"step": 10000
},
{
"epoch": 0.15,
"learning_rate": 4.227941176470588e-05,
"loss": 2.9109,
"step": 10500
},
{
"epoch": 0.16,
"learning_rate": 4.1911764705882356e-05,
"loss": 2.8797,
"step": 11000
},
{
"epoch": 0.17,
"learning_rate": 4.154411764705883e-05,
"loss": 2.8613,
"step": 11500
},
{
"epoch": 0.18,
"learning_rate": 4.11764705882353e-05,
"loss": 2.805,
"step": 12000
},
{
"epoch": 0.18,
"learning_rate": 4.0808823529411765e-05,
"loss": 2.8115,
"step": 12500
},
{
"epoch": 0.19,
"learning_rate": 4.044117647058824e-05,
"loss": 2.7802,
"step": 13000
},
{
"epoch": 0.2,
"learning_rate": 4.007352941176471e-05,
"loss": 2.7532,
"step": 13500
},
{
"epoch": 0.21,
"learning_rate": 3.970588235294117e-05,
"loss": 2.7617,
"step": 14000
},
{
"epoch": 0.21,
"learning_rate": 3.933823529411765e-05,
"loss": 2.6741,
"step": 14500
},
{
"epoch": 0.22,
"learning_rate": 3.897058823529412e-05,
"loss": 2.7255,
"step": 15000
},
{
"epoch": 0.23,
"learning_rate": 3.8602941176470595e-05,
"loss": 2.7633,
"step": 15500
},
{
"epoch": 0.24,
"learning_rate": 3.8235294117647055e-05,
"loss": 2.8108,
"step": 16000
},
{
"epoch": 0.24,
"learning_rate": 3.786764705882353e-05,
"loss": 2.7793,
"step": 16500
},
{
"epoch": 0.25,
"learning_rate": 3.7500000000000003e-05,
"loss": 2.7585,
"step": 17000
},
{
"epoch": 0.26,
"learning_rate": 3.713235294117647e-05,
"loss": 2.7341,
"step": 17500
},
{
"epoch": 0.26,
"learning_rate": 3.6764705882352945e-05,
"loss": 2.7373,
"step": 18000
},
{
"epoch": 0.27,
"learning_rate": 3.639705882352941e-05,
"loss": 2.6983,
"step": 18500
},
{
"epoch": 0.28,
"learning_rate": 3.6029411764705886e-05,
"loss": 2.7178,
"step": 19000
},
{
"epoch": 0.29,
"learning_rate": 3.566176470588235e-05,
"loss": 2.6875,
"step": 19500
},
{
"epoch": 0.29,
"learning_rate": 3.529411764705883e-05,
"loss": 2.6474,
"step": 20000
},
{
"epoch": 0.3,
"learning_rate": 3.4926470588235294e-05,
"loss": 2.6301,
"step": 20500
},
{
"epoch": 0.31,
"learning_rate": 3.455882352941177e-05,
"loss": 2.5974,
"step": 21000
},
{
"epoch": 0.32,
"learning_rate": 3.4191176470588236e-05,
"loss": 2.5881,
"step": 21500
},
{
"epoch": 0.32,
"learning_rate": 3.382352941176471e-05,
"loss": 2.5867,
"step": 22000
},
{
"epoch": 0.33,
"learning_rate": 3.345588235294118e-05,
"loss": 2.5335,
"step": 22500
},
{
"epoch": 0.34,
"learning_rate": 3.308823529411765e-05,
"loss": 2.5615,
"step": 23000
},
{
"epoch": 0.35,
"learning_rate": 3.272058823529412e-05,
"loss": 2.5329,
"step": 23500
},
{
"epoch": 0.35,
"learning_rate": 3.235294117647059e-05,
"loss": 2.5688,
"step": 24000
},
{
"epoch": 0.36,
"learning_rate": 3.198529411764706e-05,
"loss": 2.548,
"step": 24500
},
{
"epoch": 0.37,
"learning_rate": 3.161764705882353e-05,
"loss": 2.5402,
"step": 25000
},
{
"epoch": 0.38,
"learning_rate": 3.125e-05,
"loss": 2.5363,
"step": 25500
},
{
"epoch": 0.38,
"learning_rate": 3.0882352941176475e-05,
"loss": 2.5336,
"step": 26000
},
{
"epoch": 0.39,
"learning_rate": 3.0514705882352945e-05,
"loss": 2.3811,
"step": 26500
},
{
"epoch": 0.4,
"learning_rate": 3.0147058823529413e-05,
"loss": 2.2274,
"step": 27000
},
{
"epoch": 0.4,
"learning_rate": 2.9779411764705883e-05,
"loss": 2.1674,
"step": 27500
},
{
"epoch": 0.41,
"learning_rate": 2.9411764705882354e-05,
"loss": 2.2378,
"step": 28000
},
{
"epoch": 0.42,
"learning_rate": 2.9044117647058828e-05,
"loss": 2.0918,
"step": 28500
},
{
"epoch": 0.43,
"learning_rate": 2.8676470588235295e-05,
"loss": 2.1793,
"step": 29000
},
{
"epoch": 0.43,
"learning_rate": 2.8308823529411766e-05,
"loss": 2.1555,
"step": 29500
},
{
"epoch": 0.44,
"learning_rate": 2.7941176470588236e-05,
"loss": 2.1583,
"step": 30000
},
{
"epoch": 0.45,
"learning_rate": 2.757352941176471e-05,
"loss": 2.1386,
"step": 30500
},
{
"epoch": 0.46,
"learning_rate": 2.7205882352941174e-05,
"loss": 2.1928,
"step": 31000
},
{
"epoch": 0.46,
"learning_rate": 2.6838235294117648e-05,
"loss": 2.1795,
"step": 31500
},
{
"epoch": 0.47,
"learning_rate": 2.647058823529412e-05,
"loss": 2.1575,
"step": 32000
},
{
"epoch": 0.48,
"learning_rate": 2.6102941176470593e-05,
"loss": 2.1505,
"step": 32500
},
{
"epoch": 0.49,
"learning_rate": 2.5735294117647057e-05,
"loss": 2.1241,
"step": 33000
},
{
"epoch": 0.49,
"learning_rate": 2.536764705882353e-05,
"loss": 2.204,
"step": 33500
},
{
"epoch": 0.5,
"learning_rate": 2.5e-05,
"loss": 2.1518,
"step": 34000
},
{
"epoch": 0.51,
"learning_rate": 2.4632352941176472e-05,
"loss": 2.1092,
"step": 34500
},
{
"epoch": 0.51,
"learning_rate": 2.4264705882352942e-05,
"loss": 2.1507,
"step": 35000
},
{
"epoch": 0.52,
"learning_rate": 2.3897058823529413e-05,
"loss": 2.1249,
"step": 35500
},
{
"epoch": 0.53,
"learning_rate": 2.3529411764705884e-05,
"loss": 2.1248,
"step": 36000
},
{
"epoch": 0.54,
"learning_rate": 2.3161764705882354e-05,
"loss": 2.1469,
"step": 36500
},
{
"epoch": 0.54,
"learning_rate": 2.2794117647058825e-05,
"loss": 2.1257,
"step": 37000
},
{
"epoch": 0.55,
"learning_rate": 2.2426470588235296e-05,
"loss": 2.0836,
"step": 37500
},
{
"epoch": 0.56,
"learning_rate": 2.2058823529411766e-05,
"loss": 2.0994,
"step": 38000
},
{
"epoch": 0.57,
"learning_rate": 2.1691176470588237e-05,
"loss": 2.1004,
"step": 38500
},
{
"epoch": 0.57,
"learning_rate": 2.1323529411764707e-05,
"loss": 2.0787,
"step": 39000
},
{
"epoch": 0.58,
"learning_rate": 2.0955882352941178e-05,
"loss": 2.0843,
"step": 39500
},
{
"epoch": 0.59,
"learning_rate": 2.058823529411765e-05,
"loss": 2.0934,
"step": 40000
},
{
"epoch": 0.6,
"learning_rate": 2.022058823529412e-05,
"loss": 2.1403,
"step": 40500
},
{
"epoch": 0.6,
"learning_rate": 1.9852941176470586e-05,
"loss": 2.0841,
"step": 41000
},
{
"epoch": 0.61,
"learning_rate": 1.948529411764706e-05,
"loss": 2.0083,
"step": 41500
},
{
"epoch": 0.62,
"learning_rate": 1.9117647058823528e-05,
"loss": 2.1298,
"step": 42000
},
{
"epoch": 0.62,
"learning_rate": 1.8750000000000002e-05,
"loss": 2.0865,
"step": 42500
},
{
"epoch": 0.63,
"learning_rate": 1.8382352941176472e-05,
"loss": 2.0812,
"step": 43000
},
{
"epoch": 0.64,
"learning_rate": 1.8014705882352943e-05,
"loss": 2.0212,
"step": 43500
},
{
"epoch": 0.65,
"learning_rate": 1.7647058823529414e-05,
"loss": 2.0754,
"step": 44000
},
{
"epoch": 0.65,
"learning_rate": 1.7279411764705884e-05,
"loss": 2.0545,
"step": 44500
},
{
"epoch": 0.66,
"learning_rate": 1.6911764705882355e-05,
"loss": 2.0541,
"step": 45000
},
{
"epoch": 0.67,
"learning_rate": 1.6544117647058825e-05,
"loss": 2.0968,
"step": 45500
},
{
"epoch": 0.68,
"learning_rate": 1.6176470588235296e-05,
"loss": 2.0811,
"step": 46000
},
{
"epoch": 0.68,
"learning_rate": 1.5808823529411763e-05,
"loss": 2.0595,
"step": 46500
},
{
"epoch": 0.69,
"learning_rate": 1.5441176470588237e-05,
"loss": 2.0083,
"step": 47000
},
{
"epoch": 0.7,
"learning_rate": 1.5073529411764706e-05,
"loss": 1.9903,
"step": 47500
},
{
"epoch": 0.71,
"learning_rate": 1.4705882352941177e-05,
"loss": 2.0358,
"step": 48000
},
{
"epoch": 0.71,
"learning_rate": 1.4338235294117647e-05,
"loss": 2.0444,
"step": 48500
},
{
"epoch": 0.72,
"learning_rate": 1.3970588235294118e-05,
"loss": 1.9886,
"step": 49000
},
{
"epoch": 0.73,
"learning_rate": 1.3602941176470587e-05,
"loss": 2.0131,
"step": 49500
},
{
"epoch": 0.74,
"learning_rate": 1.323529411764706e-05,
"loss": 1.9727,
"step": 50000
},
{
"epoch": 0.74,
"learning_rate": 1.2867647058823528e-05,
"loss": 2.0167,
"step": 50500
},
{
"epoch": 0.75,
"learning_rate": 1.25e-05,
"loss": 2.3871,
"step": 51000
},
{
"epoch": 0.76,
"learning_rate": 1.2132352941176471e-05,
"loss": 2.6125,
"step": 51500
},
{
"epoch": 0.76,
"learning_rate": 1.1764705882352942e-05,
"loss": 2.6199,
"step": 52000
},
{
"epoch": 0.77,
"learning_rate": 1.1397058823529412e-05,
"loss": 2.6208,
"step": 52500
},
{
"epoch": 0.78,
"learning_rate": 1.1029411764705883e-05,
"loss": 2.595,
"step": 53000
},
{
"epoch": 0.79,
"learning_rate": 1.0661764705882354e-05,
"loss": 2.5738,
"step": 53500
},
{
"epoch": 0.79,
"learning_rate": 1.0294117647058824e-05,
"loss": 2.5433,
"step": 54000
},
{
"epoch": 0.8,
"learning_rate": 9.926470588235293e-06,
"loss": 2.546,
"step": 54500
},
{
"epoch": 0.81,
"learning_rate": 9.558823529411764e-06,
"loss": 2.5337,
"step": 55000
},
{
"epoch": 0.82,
"learning_rate": 9.191176470588236e-06,
"loss": 2.5338,
"step": 55500
},
{
"epoch": 0.82,
"learning_rate": 8.823529411764707e-06,
"loss": 2.4659,
"step": 56000
},
{
"epoch": 0.83,
"learning_rate": 8.455882352941177e-06,
"loss": 2.4391,
"step": 56500
},
{
"epoch": 0.84,
"learning_rate": 8.088235294117648e-06,
"loss": 2.4408,
"step": 57000
},
{
"epoch": 0.85,
"learning_rate": 7.720588235294119e-06,
"loss": 2.4144,
"step": 57500
},
{
"epoch": 0.85,
"learning_rate": 7.3529411764705884e-06,
"loss": 2.4124,
"step": 58000
},
{
"epoch": 0.86,
"learning_rate": 6.985294117647059e-06,
"loss": 2.4135,
"step": 58500
},
{
"epoch": 0.87,
"learning_rate": 6.61764705882353e-06,
"loss": 2.4227,
"step": 59000
},
{
"epoch": 0.88,
"learning_rate": 6.25e-06,
"loss": 2.4051,
"step": 59500
},
{
"epoch": 0.88,
"learning_rate": 5.882352941176471e-06,
"loss": 2.4224,
"step": 60000
},
{
"epoch": 0.89,
"learning_rate": 5.5147058823529415e-06,
"loss": 2.3938,
"step": 60500
},
{
"epoch": 0.9,
"learning_rate": 5.147058823529412e-06,
"loss": 2.4126,
"step": 61000
},
{
"epoch": 0.9,
"learning_rate": 4.779411764705882e-06,
"loss": 2.4084,
"step": 61500
},
{
"epoch": 0.91,
"learning_rate": 4.411764705882353e-06,
"loss": 2.4141,
"step": 62000
},
{
"epoch": 0.92,
"learning_rate": 4.044117647058824e-06,
"loss": 2.4128,
"step": 62500
},
{
"epoch": 0.93,
"learning_rate": 3.6764705882352942e-06,
"loss": 2.381,
"step": 63000
},
{
"epoch": 0.93,
"learning_rate": 3.308823529411765e-06,
"loss": 2.3949,
"step": 63500
},
{
"epoch": 0.94,
"learning_rate": 2.9411764705882355e-06,
"loss": 2.3984,
"step": 64000
},
{
"epoch": 0.95,
"learning_rate": 2.573529411764706e-06,
"loss": 2.3848,
"step": 64500
},
{
"epoch": 0.96,
"learning_rate": 2.2058823529411767e-06,
"loss": 2.4009,
"step": 65000
},
{
"epoch": 0.96,
"learning_rate": 1.8382352941176471e-06,
"loss": 2.3544,
"step": 65500
},
{
"epoch": 0.97,
"learning_rate": 1.4705882352941177e-06,
"loss": 2.3832,
"step": 66000
},
{
"epoch": 0.98,
"learning_rate": 1.1029411764705884e-06,
"loss": 2.3815,
"step": 66500
},
{
"epoch": 0.99,
"learning_rate": 7.352941176470589e-07,
"loss": 2.6238,
"step": 67000
},
{
"epoch": 0.99,
"learning_rate": 3.6764705882352943e-07,
"loss": 2.5585,
"step": 67500
},
{
"epoch": 1.0,
"learning_rate": 0.0,
"loss": 2.482,
"step": 68000
}
],
"logging_steps": 500,
"max_steps": 68000,
"num_train_epochs": 9223372036854775807,
"save_steps": 4000,
"total_flos": 4.041705422585856e+18,
"trial_name": null,
"trial_params": null
}