{ "best_metric": 0.02364550158381462, "best_model_checkpoint": "/storage/img-cls-data/mnist_outputs/checkpoint-31875", "epoch": 5.0, "global_step": 31875, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.02, "learning_rate": 1.9937254901960785e-05, "loss": 2.0433, "step": 100 }, { "epoch": 0.03, "learning_rate": 1.987450980392157e-05, "loss": 1.5446, "step": 200 }, { "epoch": 0.05, "learning_rate": 1.9811764705882354e-05, "loss": 1.2233, "step": 300 }, { "epoch": 0.06, "learning_rate": 1.9749019607843137e-05, "loss": 1.0046, "step": 400 }, { "epoch": 0.08, "learning_rate": 1.9686274509803924e-05, "loss": 0.9327, "step": 500 }, { "epoch": 0.09, "learning_rate": 1.9623529411764707e-05, "loss": 0.8659, "step": 600 }, { "epoch": 0.11, "learning_rate": 1.9560784313725493e-05, "loss": 0.7593, "step": 700 }, { "epoch": 0.13, "learning_rate": 1.9498039215686276e-05, "loss": 0.7057, "step": 800 }, { "epoch": 0.14, "learning_rate": 1.9435294117647063e-05, "loss": 0.6666, "step": 900 }, { "epoch": 0.16, "learning_rate": 1.9372549019607846e-05, "loss": 0.6416, "step": 1000 }, { "epoch": 0.17, "learning_rate": 1.930980392156863e-05, "loss": 0.665, "step": 1100 }, { "epoch": 0.19, "learning_rate": 1.9247058823529415e-05, "loss": 0.6084, "step": 1200 }, { "epoch": 0.2, "learning_rate": 1.9184313725490198e-05, "loss": 0.6203, "step": 1300 }, { "epoch": 0.22, "learning_rate": 1.912156862745098e-05, "loss": 0.6202, "step": 1400 }, { "epoch": 0.24, "learning_rate": 1.9058823529411764e-05, "loss": 0.5994, "step": 1500 }, { "epoch": 0.25, "learning_rate": 1.899607843137255e-05, "loss": 0.5837, "step": 1600 }, { "epoch": 0.27, "learning_rate": 1.8933333333333334e-05, "loss": 0.5787, "step": 1700 }, { "epoch": 0.28, "learning_rate": 1.8870588235294117e-05, "loss": 0.573, "step": 1800 }, { "epoch": 0.3, "learning_rate": 1.8807843137254903e-05, "loss": 0.5547, "step": 1900 }, { "epoch": 0.31, "learning_rate": 1.8745098039215686e-05, "loss": 0.5362, "step": 2000 }, { "epoch": 0.33, "learning_rate": 1.8682352941176473e-05, "loss": 0.5462, "step": 2100 }, { "epoch": 0.35, "learning_rate": 1.8619607843137256e-05, "loss": 0.4988, "step": 2200 }, { "epoch": 0.36, "learning_rate": 1.8556862745098042e-05, "loss": 0.4175, "step": 2300 }, { "epoch": 0.38, "learning_rate": 1.8494117647058825e-05, "loss": 0.5065, "step": 2400 }, { "epoch": 0.39, "learning_rate": 1.843137254901961e-05, "loss": 0.5204, "step": 2500 }, { "epoch": 0.41, "learning_rate": 1.8368627450980395e-05, "loss": 0.4434, "step": 2600 }, { "epoch": 0.42, "learning_rate": 1.8305882352941178e-05, "loss": 0.5012, "step": 2700 }, { "epoch": 0.44, "learning_rate": 1.824313725490196e-05, "loss": 0.5101, "step": 2800 }, { "epoch": 0.45, "learning_rate": 1.8180392156862747e-05, "loss": 0.4713, "step": 2900 }, { "epoch": 0.47, "learning_rate": 1.811764705882353e-05, "loss": 0.4327, "step": 3000 }, { "epoch": 0.49, "learning_rate": 1.8054901960784314e-05, "loss": 0.5051, "step": 3100 }, { "epoch": 0.5, "learning_rate": 1.79921568627451e-05, "loss": 0.4257, "step": 3200 }, { "epoch": 0.52, "learning_rate": 1.7929411764705883e-05, "loss": 0.427, "step": 3300 }, { "epoch": 0.53, "learning_rate": 1.7866666666666666e-05, "loss": 0.4886, "step": 3400 }, { "epoch": 0.55, "learning_rate": 1.7803921568627453e-05, "loss": 0.4276, "step": 3500 }, { "epoch": 0.56, "learning_rate": 1.7741176470588236e-05, "loss": 0.4275, "step": 3600 }, { "epoch": 0.58, "learning_rate": 1.7678431372549022e-05, "loss": 0.4558, "step": 3700 }, { "epoch": 0.6, "learning_rate": 1.7615686274509805e-05, "loss": 0.4459, "step": 3800 }, { "epoch": 0.61, "learning_rate": 1.755294117647059e-05, "loss": 0.5111, "step": 3900 }, { "epoch": 0.63, "learning_rate": 1.7490196078431375e-05, "loss": 0.3821, "step": 4000 }, { "epoch": 0.64, "learning_rate": 1.7427450980392158e-05, "loss": 0.4291, "step": 4100 }, { "epoch": 0.66, "learning_rate": 1.7364705882352944e-05, "loss": 0.4001, "step": 4200 }, { "epoch": 0.67, "learning_rate": 1.7301960784313727e-05, "loss": 0.4144, "step": 4300 }, { "epoch": 0.69, "learning_rate": 1.723921568627451e-05, "loss": 0.3786, "step": 4400 }, { "epoch": 0.71, "learning_rate": 1.7176470588235293e-05, "loss": 0.4464, "step": 4500 }, { "epoch": 0.72, "learning_rate": 1.711372549019608e-05, "loss": 0.449, "step": 4600 }, { "epoch": 0.74, "learning_rate": 1.7050980392156863e-05, "loss": 0.4206, "step": 4700 }, { "epoch": 0.75, "learning_rate": 1.698823529411765e-05, "loss": 0.4625, "step": 4800 }, { "epoch": 0.77, "learning_rate": 1.6925490196078432e-05, "loss": 0.4383, "step": 4900 }, { "epoch": 0.78, "learning_rate": 1.686274509803922e-05, "loss": 0.3844, "step": 5000 }, { "epoch": 0.8, "learning_rate": 1.6800000000000002e-05, "loss": 0.4103, "step": 5100 }, { "epoch": 0.82, "learning_rate": 1.6737254901960788e-05, "loss": 0.4965, "step": 5200 }, { "epoch": 0.83, "learning_rate": 1.667450980392157e-05, "loss": 0.368, "step": 5300 }, { "epoch": 0.85, "learning_rate": 1.6611764705882354e-05, "loss": 0.4474, "step": 5400 }, { "epoch": 0.86, "learning_rate": 1.6549019607843137e-05, "loss": 0.4173, "step": 5500 }, { "epoch": 0.88, "learning_rate": 1.6486274509803924e-05, "loss": 0.4226, "step": 5600 }, { "epoch": 0.89, "learning_rate": 1.6423529411764707e-05, "loss": 0.4163, "step": 5700 }, { "epoch": 0.91, "learning_rate": 1.636078431372549e-05, "loss": 0.4096, "step": 5800 }, { "epoch": 0.93, "learning_rate": 1.6298039215686276e-05, "loss": 0.3749, "step": 5900 }, { "epoch": 0.94, "learning_rate": 1.623529411764706e-05, "loss": 0.4418, "step": 6000 }, { "epoch": 0.96, "learning_rate": 1.6172549019607842e-05, "loss": 0.4591, "step": 6100 }, { "epoch": 0.97, "learning_rate": 1.610980392156863e-05, "loss": 0.4181, "step": 6200 }, { "epoch": 0.99, "learning_rate": 1.6047058823529412e-05, "loss": 0.3717, "step": 6300 }, { "epoch": 1.0, "eval_accuracy": 0.9893333333333333, "eval_loss": 0.05222630500793457, "eval_runtime": 45.5882, "eval_samples_per_second": 197.42, "eval_steps_per_second": 24.677, "step": 6375 }, { "epoch": 1.0, "learning_rate": 1.59843137254902e-05, "loss": 0.3852, "step": 6400 }, { "epoch": 1.02, "learning_rate": 1.592156862745098e-05, "loss": 0.4254, "step": 6500 }, { "epoch": 1.04, "learning_rate": 1.5858823529411768e-05, "loss": 0.4035, "step": 6600 }, { "epoch": 1.05, "learning_rate": 1.579607843137255e-05, "loss": 0.3704, "step": 6700 }, { "epoch": 1.07, "learning_rate": 1.5733333333333334e-05, "loss": 0.4037, "step": 6800 }, { "epoch": 1.08, "learning_rate": 1.567058823529412e-05, "loss": 0.3574, "step": 6900 }, { "epoch": 1.1, "learning_rate": 1.5607843137254904e-05, "loss": 0.3766, "step": 7000 }, { "epoch": 1.11, "learning_rate": 1.5545098039215687e-05, "loss": 0.3677, "step": 7100 }, { "epoch": 1.13, "learning_rate": 1.5482352941176473e-05, "loss": 0.5049, "step": 7200 }, { "epoch": 1.15, "learning_rate": 1.5419607843137256e-05, "loss": 0.3507, "step": 7300 }, { "epoch": 1.16, "learning_rate": 1.535686274509804e-05, "loss": 0.4128, "step": 7400 }, { "epoch": 1.18, "learning_rate": 1.5294117647058822e-05, "loss": 0.3476, "step": 7500 }, { "epoch": 1.19, "learning_rate": 1.523137254901961e-05, "loss": 0.3652, "step": 7600 }, { "epoch": 1.21, "learning_rate": 1.5168627450980393e-05, "loss": 0.3738, "step": 7700 }, { "epoch": 1.22, "learning_rate": 1.5105882352941176e-05, "loss": 0.3927, "step": 7800 }, { "epoch": 1.24, "learning_rate": 1.5043137254901963e-05, "loss": 0.3556, "step": 7900 }, { "epoch": 1.25, "learning_rate": 1.4980392156862746e-05, "loss": 0.3917, "step": 8000 }, { "epoch": 1.27, "learning_rate": 1.491764705882353e-05, "loss": 0.3965, "step": 8100 }, { "epoch": 1.29, "learning_rate": 1.4854901960784315e-05, "loss": 0.3356, "step": 8200 }, { "epoch": 1.3, "learning_rate": 1.47921568627451e-05, "loss": 0.3583, "step": 8300 }, { "epoch": 1.32, "learning_rate": 1.4729411764705883e-05, "loss": 0.4065, "step": 8400 }, { "epoch": 1.33, "learning_rate": 1.4666666666666666e-05, "loss": 0.4424, "step": 8500 }, { "epoch": 1.35, "learning_rate": 1.4603921568627453e-05, "loss": 0.3177, "step": 8600 }, { "epoch": 1.36, "learning_rate": 1.4541176470588236e-05, "loss": 0.3676, "step": 8700 }, { "epoch": 1.38, "learning_rate": 1.447843137254902e-05, "loss": 0.3968, "step": 8800 }, { "epoch": 1.4, "learning_rate": 1.4415686274509805e-05, "loss": 0.3891, "step": 8900 }, { "epoch": 1.41, "learning_rate": 1.435294117647059e-05, "loss": 0.4299, "step": 9000 }, { "epoch": 1.43, "learning_rate": 1.4290196078431373e-05, "loss": 0.3649, "step": 9100 }, { "epoch": 1.44, "learning_rate": 1.422745098039216e-05, "loss": 0.3936, "step": 9200 }, { "epoch": 1.46, "learning_rate": 1.4164705882352943e-05, "loss": 0.3488, "step": 9300 }, { "epoch": 1.47, "learning_rate": 1.4101960784313726e-05, "loss": 0.3544, "step": 9400 }, { "epoch": 1.49, "learning_rate": 1.403921568627451e-05, "loss": 0.3886, "step": 9500 }, { "epoch": 1.51, "learning_rate": 1.3976470588235295e-05, "loss": 0.3687, "step": 9600 }, { "epoch": 1.52, "learning_rate": 1.391372549019608e-05, "loss": 0.3756, "step": 9700 }, { "epoch": 1.54, "learning_rate": 1.3850980392156863e-05, "loss": 0.3966, "step": 9800 }, { "epoch": 1.55, "learning_rate": 1.378823529411765e-05, "loss": 0.3965, "step": 9900 }, { "epoch": 1.57, "learning_rate": 1.3725490196078432e-05, "loss": 0.4041, "step": 10000 }, { "epoch": 1.58, "learning_rate": 1.3662745098039215e-05, "loss": 0.321, "step": 10100 }, { "epoch": 1.6, "learning_rate": 1.3600000000000002e-05, "loss": 0.34, "step": 10200 }, { "epoch": 1.62, "learning_rate": 1.3537254901960785e-05, "loss": 0.3859, "step": 10300 }, { "epoch": 1.63, "learning_rate": 1.347450980392157e-05, "loss": 0.3574, "step": 10400 }, { "epoch": 1.65, "learning_rate": 1.3411764705882353e-05, "loss": 0.3645, "step": 10500 }, { "epoch": 1.66, "learning_rate": 1.334901960784314e-05, "loss": 0.3378, "step": 10600 }, { "epoch": 1.68, "learning_rate": 1.3286274509803922e-05, "loss": 0.4109, "step": 10700 }, { "epoch": 1.69, "learning_rate": 1.3223529411764705e-05, "loss": 0.3732, "step": 10800 }, { "epoch": 1.71, "learning_rate": 1.3160784313725492e-05, "loss": 0.3548, "step": 10900 }, { "epoch": 1.73, "learning_rate": 1.3098039215686275e-05, "loss": 0.3076, "step": 11000 }, { "epoch": 1.74, "learning_rate": 1.303529411764706e-05, "loss": 0.3707, "step": 11100 }, { "epoch": 1.76, "learning_rate": 1.2972549019607846e-05, "loss": 0.367, "step": 11200 }, { "epoch": 1.77, "learning_rate": 1.2909803921568629e-05, "loss": 0.3242, "step": 11300 }, { "epoch": 1.79, "learning_rate": 1.2847058823529412e-05, "loss": 0.3913, "step": 11400 }, { "epoch": 1.8, "learning_rate": 1.2784313725490197e-05, "loss": 0.3605, "step": 11500 }, { "epoch": 1.82, "learning_rate": 1.2721568627450982e-05, "loss": 0.352, "step": 11600 }, { "epoch": 1.84, "learning_rate": 1.2658823529411766e-05, "loss": 0.3326, "step": 11700 }, { "epoch": 1.85, "learning_rate": 1.259607843137255e-05, "loss": 0.3601, "step": 11800 }, { "epoch": 1.87, "learning_rate": 1.2533333333333336e-05, "loss": 0.389, "step": 11900 }, { "epoch": 1.88, "learning_rate": 1.2470588235294119e-05, "loss": 0.3516, "step": 12000 }, { "epoch": 1.9, "learning_rate": 1.2407843137254902e-05, "loss": 0.3759, "step": 12100 }, { "epoch": 1.91, "learning_rate": 1.2345098039215688e-05, "loss": 0.3405, "step": 12200 }, { "epoch": 1.93, "learning_rate": 1.2282352941176471e-05, "loss": 0.3621, "step": 12300 }, { "epoch": 1.95, "learning_rate": 1.2219607843137256e-05, "loss": 0.3429, "step": 12400 }, { "epoch": 1.96, "learning_rate": 1.215686274509804e-05, "loss": 0.3789, "step": 12500 }, { "epoch": 1.98, "learning_rate": 1.2094117647058826e-05, "loss": 0.3011, "step": 12600 }, { "epoch": 1.99, "learning_rate": 1.2031372549019609e-05, "loss": 0.3453, "step": 12700 }, { "epoch": 2.0, "eval_accuracy": 0.9905555555555555, "eval_loss": 0.0370011106133461, "eval_runtime": 47.336, "eval_samples_per_second": 190.13, "eval_steps_per_second": 23.766, "step": 12750 }, { "epoch": 2.01, "learning_rate": 1.1968627450980392e-05, "loss": 0.3666, "step": 12800 }, { "epoch": 2.02, "learning_rate": 1.1905882352941178e-05, "loss": 0.3447, "step": 12900 }, { "epoch": 2.04, "learning_rate": 1.1843137254901961e-05, "loss": 0.3669, "step": 13000 }, { "epoch": 2.05, "learning_rate": 1.1780392156862746e-05, "loss": 0.3423, "step": 13100 }, { "epoch": 2.07, "learning_rate": 1.171764705882353e-05, "loss": 0.3372, "step": 13200 }, { "epoch": 2.09, "learning_rate": 1.1654901960784316e-05, "loss": 0.356, "step": 13300 }, { "epoch": 2.1, "learning_rate": 1.1592156862745099e-05, "loss": 0.306, "step": 13400 }, { "epoch": 2.12, "learning_rate": 1.1529411764705882e-05, "loss": 0.354, "step": 13500 }, { "epoch": 2.13, "learning_rate": 1.1466666666666668e-05, "loss": 0.3664, "step": 13600 }, { "epoch": 2.15, "learning_rate": 1.1403921568627451e-05, "loss": 0.3568, "step": 13700 }, { "epoch": 2.16, "learning_rate": 1.1341176470588236e-05, "loss": 0.3334, "step": 13800 }, { "epoch": 2.18, "learning_rate": 1.127843137254902e-05, "loss": 0.3405, "step": 13900 }, { "epoch": 2.2, "learning_rate": 1.1215686274509805e-05, "loss": 0.3107, "step": 14000 }, { "epoch": 2.21, "learning_rate": 1.1152941176470588e-05, "loss": 0.2895, "step": 14100 }, { "epoch": 2.23, "learning_rate": 1.1090196078431375e-05, "loss": 0.2832, "step": 14200 }, { "epoch": 2.24, "learning_rate": 1.1027450980392158e-05, "loss": 0.3823, "step": 14300 }, { "epoch": 2.26, "learning_rate": 1.0964705882352941e-05, "loss": 0.2835, "step": 14400 }, { "epoch": 2.27, "learning_rate": 1.0901960784313726e-05, "loss": 0.3859, "step": 14500 }, { "epoch": 2.29, "learning_rate": 1.083921568627451e-05, "loss": 0.3486, "step": 14600 }, { "epoch": 2.31, "learning_rate": 1.0776470588235295e-05, "loss": 0.3202, "step": 14700 }, { "epoch": 2.32, "learning_rate": 1.0713725490196078e-05, "loss": 0.3276, "step": 14800 }, { "epoch": 2.34, "learning_rate": 1.0650980392156865e-05, "loss": 0.3562, "step": 14900 }, { "epoch": 2.35, "learning_rate": 1.0588235294117648e-05, "loss": 0.3344, "step": 15000 }, { "epoch": 2.37, "learning_rate": 1.0525490196078431e-05, "loss": 0.3356, "step": 15100 }, { "epoch": 2.38, "learning_rate": 1.0462745098039217e-05, "loss": 0.322, "step": 15200 }, { "epoch": 2.4, "learning_rate": 1.04e-05, "loss": 0.3114, "step": 15300 }, { "epoch": 2.42, "learning_rate": 1.0337254901960785e-05, "loss": 0.3401, "step": 15400 }, { "epoch": 2.43, "learning_rate": 1.0274509803921568e-05, "loss": 0.3624, "step": 15500 }, { "epoch": 2.45, "learning_rate": 1.0211764705882355e-05, "loss": 0.3468, "step": 15600 }, { "epoch": 2.46, "learning_rate": 1.0149019607843138e-05, "loss": 0.3263, "step": 15700 }, { "epoch": 2.48, "learning_rate": 1.0086274509803922e-05, "loss": 0.317, "step": 15800 }, { "epoch": 2.49, "learning_rate": 1.0023529411764707e-05, "loss": 0.4155, "step": 15900 }, { "epoch": 2.51, "learning_rate": 9.960784313725492e-06, "loss": 0.2799, "step": 16000 }, { "epoch": 2.53, "learning_rate": 9.898039215686275e-06, "loss": 0.3184, "step": 16100 }, { "epoch": 2.54, "learning_rate": 9.83529411764706e-06, "loss": 0.3319, "step": 16200 }, { "epoch": 2.56, "learning_rate": 9.772549019607844e-06, "loss": 0.339, "step": 16300 }, { "epoch": 2.57, "learning_rate": 9.709803921568628e-06, "loss": 0.3767, "step": 16400 }, { "epoch": 2.59, "learning_rate": 9.647058823529412e-06, "loss": 0.2787, "step": 16500 }, { "epoch": 2.6, "learning_rate": 9.584313725490197e-06, "loss": 0.2856, "step": 16600 }, { "epoch": 2.62, "learning_rate": 9.521568627450982e-06, "loss": 0.3085, "step": 16700 }, { "epoch": 2.64, "learning_rate": 9.458823529411767e-06, "loss": 0.3445, "step": 16800 }, { "epoch": 2.65, "learning_rate": 9.39607843137255e-06, "loss": 0.3586, "step": 16900 }, { "epoch": 2.67, "learning_rate": 9.333333333333334e-06, "loss": 0.2763, "step": 17000 }, { "epoch": 2.68, "learning_rate": 9.270588235294117e-06, "loss": 0.3879, "step": 17100 }, { "epoch": 2.7, "learning_rate": 9.207843137254902e-06, "loss": 0.3333, "step": 17200 }, { "epoch": 2.71, "learning_rate": 9.145098039215687e-06, "loss": 0.3134, "step": 17300 }, { "epoch": 2.73, "learning_rate": 9.082352941176472e-06, "loss": 0.3143, "step": 17400 }, { "epoch": 2.75, "learning_rate": 9.019607843137256e-06, "loss": 0.3189, "step": 17500 }, { "epoch": 2.76, "learning_rate": 8.95686274509804e-06, "loss": 0.3751, "step": 17600 }, { "epoch": 2.78, "learning_rate": 8.894117647058824e-06, "loss": 0.3392, "step": 17700 }, { "epoch": 2.79, "learning_rate": 8.831372549019609e-06, "loss": 0.4085, "step": 17800 }, { "epoch": 2.81, "learning_rate": 8.768627450980392e-06, "loss": 0.351, "step": 17900 }, { "epoch": 2.82, "learning_rate": 8.705882352941177e-06, "loss": 0.4239, "step": 18000 }, { "epoch": 2.84, "learning_rate": 8.643137254901961e-06, "loss": 0.3577, "step": 18100 }, { "epoch": 2.85, "learning_rate": 8.580392156862746e-06, "loss": 0.3651, "step": 18200 }, { "epoch": 2.87, "learning_rate": 8.517647058823531e-06, "loss": 0.2912, "step": 18300 }, { "epoch": 2.89, "learning_rate": 8.454901960784314e-06, "loss": 0.3367, "step": 18400 }, { "epoch": 2.9, "learning_rate": 8.392156862745099e-06, "loss": 0.3616, "step": 18500 }, { "epoch": 2.92, "learning_rate": 8.329411764705882e-06, "loss": 0.3608, "step": 18600 }, { "epoch": 2.93, "learning_rate": 8.266666666666667e-06, "loss": 0.3373, "step": 18700 }, { "epoch": 2.95, "learning_rate": 8.203921568627451e-06, "loss": 0.3132, "step": 18800 }, { "epoch": 2.96, "learning_rate": 8.141176470588236e-06, "loss": 0.3279, "step": 18900 }, { "epoch": 2.98, "learning_rate": 8.07843137254902e-06, "loss": 0.3052, "step": 19000 }, { "epoch": 3.0, "learning_rate": 8.015686274509804e-06, "loss": 0.3736, "step": 19100 }, { "epoch": 3.0, "eval_accuracy": 0.9915555555555555, "eval_loss": 0.030801517888903618, "eval_runtime": 47.8145, "eval_samples_per_second": 188.227, "eval_steps_per_second": 23.528, "step": 19125 }, { "epoch": 3.01, "learning_rate": 7.952941176470589e-06, "loss": 0.3429, "step": 19200 }, { "epoch": 3.03, "learning_rate": 7.890196078431373e-06, "loss": 0.3234, "step": 19300 }, { "epoch": 3.04, "learning_rate": 7.827450980392156e-06, "loss": 0.3176, "step": 19400 }, { "epoch": 3.06, "learning_rate": 7.764705882352941e-06, "loss": 0.3483, "step": 19500 }, { "epoch": 3.07, "learning_rate": 7.701960784313726e-06, "loss": 0.2904, "step": 19600 }, { "epoch": 3.09, "learning_rate": 7.63921568627451e-06, "loss": 0.3505, "step": 19700 }, { "epoch": 3.11, "learning_rate": 7.576470588235295e-06, "loss": 0.2773, "step": 19800 }, { "epoch": 3.12, "learning_rate": 7.5137254901960785e-06, "loss": 0.3035, "step": 19900 }, { "epoch": 3.14, "learning_rate": 7.450980392156863e-06, "loss": 0.262, "step": 20000 }, { "epoch": 3.15, "learning_rate": 7.388235294117647e-06, "loss": 0.288, "step": 20100 }, { "epoch": 3.17, "learning_rate": 7.325490196078432e-06, "loss": 0.2816, "step": 20200 }, { "epoch": 3.18, "learning_rate": 7.262745098039217e-06, "loss": 0.3427, "step": 20300 }, { "epoch": 3.2, "learning_rate": 7.2000000000000005e-06, "loss": 0.3235, "step": 20400 }, { "epoch": 3.22, "learning_rate": 7.137254901960785e-06, "loss": 0.305, "step": 20500 }, { "epoch": 3.23, "learning_rate": 7.074509803921568e-06, "loss": 0.3213, "step": 20600 }, { "epoch": 3.25, "learning_rate": 7.011764705882353e-06, "loss": 0.2998, "step": 20700 }, { "epoch": 3.26, "learning_rate": 6.949019607843138e-06, "loss": 0.2733, "step": 20800 }, { "epoch": 3.28, "learning_rate": 6.886274509803922e-06, "loss": 0.2812, "step": 20900 }, { "epoch": 3.29, "learning_rate": 6.8235294117647065e-06, "loss": 0.2979, "step": 21000 }, { "epoch": 3.31, "learning_rate": 6.76078431372549e-06, "loss": 0.3359, "step": 21100 }, { "epoch": 3.33, "learning_rate": 6.698039215686275e-06, "loss": 0.2952, "step": 21200 }, { "epoch": 3.34, "learning_rate": 6.63529411764706e-06, "loss": 0.3092, "step": 21300 }, { "epoch": 3.36, "learning_rate": 6.572549019607844e-06, "loss": 0.2614, "step": 21400 }, { "epoch": 3.37, "learning_rate": 6.5098039215686285e-06, "loss": 0.3431, "step": 21500 }, { "epoch": 3.39, "learning_rate": 6.4470588235294116e-06, "loss": 0.355, "step": 21600 }, { "epoch": 3.4, "learning_rate": 6.384313725490196e-06, "loss": 0.2789, "step": 21700 }, { "epoch": 3.42, "learning_rate": 6.321568627450981e-06, "loss": 0.3669, "step": 21800 }, { "epoch": 3.44, "learning_rate": 6.258823529411765e-06, "loss": 0.2962, "step": 21900 }, { "epoch": 3.45, "learning_rate": 6.19607843137255e-06, "loss": 0.3174, "step": 22000 }, { "epoch": 3.47, "learning_rate": 6.133333333333334e-06, "loss": 0.3251, "step": 22100 }, { "epoch": 3.48, "learning_rate": 6.070588235294118e-06, "loss": 0.3237, "step": 22200 }, { "epoch": 3.5, "learning_rate": 6.007843137254903e-06, "loss": 0.2878, "step": 22300 }, { "epoch": 3.51, "learning_rate": 5.945098039215686e-06, "loss": 0.2878, "step": 22400 }, { "epoch": 3.53, "learning_rate": 5.882352941176471e-06, "loss": 0.3223, "step": 22500 }, { "epoch": 3.55, "learning_rate": 5.819607843137255e-06, "loss": 0.3064, "step": 22600 }, { "epoch": 3.56, "learning_rate": 5.7568627450980396e-06, "loss": 0.3291, "step": 22700 }, { "epoch": 3.58, "learning_rate": 5.694117647058824e-06, "loss": 0.3049, "step": 22800 }, { "epoch": 3.59, "learning_rate": 5.631372549019608e-06, "loss": 0.3109, "step": 22900 }, { "epoch": 3.61, "learning_rate": 5.568627450980393e-06, "loss": 0.374, "step": 23000 }, { "epoch": 3.62, "learning_rate": 5.505882352941177e-06, "loss": 0.3325, "step": 23100 }, { "epoch": 3.64, "learning_rate": 5.443137254901962e-06, "loss": 0.3247, "step": 23200 }, { "epoch": 3.65, "learning_rate": 5.380392156862746e-06, "loss": 0.3483, "step": 23300 }, { "epoch": 3.67, "learning_rate": 5.317647058823529e-06, "loss": 0.3321, "step": 23400 }, { "epoch": 3.69, "learning_rate": 5.254901960784314e-06, "loss": 0.2998, "step": 23500 }, { "epoch": 3.7, "learning_rate": 5.192156862745098e-06, "loss": 0.3337, "step": 23600 }, { "epoch": 3.72, "learning_rate": 5.129411764705883e-06, "loss": 0.399, "step": 23700 }, { "epoch": 3.73, "learning_rate": 5.0666666666666676e-06, "loss": 0.2966, "step": 23800 }, { "epoch": 3.75, "learning_rate": 5.0039215686274515e-06, "loss": 0.307, "step": 23900 }, { "epoch": 3.76, "learning_rate": 4.941176470588236e-06, "loss": 0.2964, "step": 24000 }, { "epoch": 3.78, "learning_rate": 4.87843137254902e-06, "loss": 0.3022, "step": 24100 }, { "epoch": 3.8, "learning_rate": 4.815686274509804e-06, "loss": 0.2502, "step": 24200 }, { "epoch": 3.81, "learning_rate": 4.752941176470589e-06, "loss": 0.2361, "step": 24300 }, { "epoch": 3.83, "learning_rate": 4.690196078431373e-06, "loss": 0.2654, "step": 24400 }, { "epoch": 3.84, "learning_rate": 4.627450980392157e-06, "loss": 0.2892, "step": 24500 }, { "epoch": 3.86, "learning_rate": 4.564705882352941e-06, "loss": 0.3175, "step": 24600 }, { "epoch": 3.87, "learning_rate": 4.501960784313726e-06, "loss": 0.3644, "step": 24700 }, { "epoch": 3.89, "learning_rate": 4.43921568627451e-06, "loss": 0.2597, "step": 24800 }, { "epoch": 3.91, "learning_rate": 4.376470588235294e-06, "loss": 0.3063, "step": 24900 }, { "epoch": 3.92, "learning_rate": 4.313725490196079e-06, "loss": 0.3096, "step": 25000 }, { "epoch": 3.94, "learning_rate": 4.250980392156863e-06, "loss": 0.3088, "step": 25100 }, { "epoch": 3.95, "learning_rate": 4.188235294117647e-06, "loss": 0.3462, "step": 25200 }, { "epoch": 3.97, "learning_rate": 4.125490196078432e-06, "loss": 0.2733, "step": 25300 }, { "epoch": 3.98, "learning_rate": 4.062745098039216e-06, "loss": 0.2905, "step": 25400 }, { "epoch": 4.0, "learning_rate": 4.000000000000001e-06, "loss": 0.3224, "step": 25500 }, { "epoch": 4.0, "eval_accuracy": 0.9938888888888889, "eval_loss": 0.026916783303022385, "eval_runtime": 48.3846, "eval_samples_per_second": 186.009, "eval_steps_per_second": 23.251, "step": 25500 }, { "epoch": 4.02, "learning_rate": 3.9372549019607846e-06, "loss": 0.2968, "step": 25600 }, { "epoch": 4.03, "learning_rate": 3.874509803921569e-06, "loss": 0.238, "step": 25700 }, { "epoch": 4.05, "learning_rate": 3.8117647058823532e-06, "loss": 0.2669, "step": 25800 }, { "epoch": 4.06, "learning_rate": 3.7490196078431375e-06, "loss": 0.2983, "step": 25900 }, { "epoch": 4.08, "learning_rate": 3.6862745098039223e-06, "loss": 0.2686, "step": 26000 }, { "epoch": 4.09, "learning_rate": 3.623529411764706e-06, "loss": 0.2907, "step": 26100 }, { "epoch": 4.11, "learning_rate": 3.5607843137254905e-06, "loss": 0.3151, "step": 26200 }, { "epoch": 4.13, "learning_rate": 3.498039215686275e-06, "loss": 0.2993, "step": 26300 }, { "epoch": 4.14, "learning_rate": 3.4352941176470587e-06, "loss": 0.2732, "step": 26400 }, { "epoch": 4.16, "learning_rate": 3.3725490196078435e-06, "loss": 0.2435, "step": 26500 }, { "epoch": 4.17, "learning_rate": 3.309803921568628e-06, "loss": 0.3121, "step": 26600 }, { "epoch": 4.19, "learning_rate": 3.247058823529412e-06, "loss": 0.2927, "step": 26700 }, { "epoch": 4.2, "learning_rate": 3.184313725490196e-06, "loss": 0.3005, "step": 26800 }, { "epoch": 4.22, "learning_rate": 3.1215686274509804e-06, "loss": 0.286, "step": 26900 }, { "epoch": 4.24, "learning_rate": 3.058823529411765e-06, "loss": 0.2461, "step": 27000 }, { "epoch": 4.25, "learning_rate": 2.9960784313725494e-06, "loss": 0.2951, "step": 27100 }, { "epoch": 4.27, "learning_rate": 2.9333333333333338e-06, "loss": 0.2699, "step": 27200 }, { "epoch": 4.28, "learning_rate": 2.8705882352941177e-06, "loss": 0.3023, "step": 27300 }, { "epoch": 4.3, "learning_rate": 2.807843137254902e-06, "loss": 0.2508, "step": 27400 }, { "epoch": 4.31, "learning_rate": 2.7450980392156867e-06, "loss": 0.2879, "step": 27500 }, { "epoch": 4.33, "learning_rate": 2.682352941176471e-06, "loss": 0.3172, "step": 27600 }, { "epoch": 4.35, "learning_rate": 2.619607843137255e-06, "loss": 0.3033, "step": 27700 }, { "epoch": 4.36, "learning_rate": 2.5568627450980393e-06, "loss": 0.3222, "step": 27800 }, { "epoch": 4.38, "learning_rate": 2.4941176470588236e-06, "loss": 0.373, "step": 27900 }, { "epoch": 4.39, "learning_rate": 2.431372549019608e-06, "loss": 0.3396, "step": 28000 }, { "epoch": 4.41, "learning_rate": 2.3686274509803927e-06, "loss": 0.2794, "step": 28100 }, { "epoch": 4.42, "learning_rate": 2.3058823529411766e-06, "loss": 0.3277, "step": 28200 }, { "epoch": 4.44, "learning_rate": 2.243137254901961e-06, "loss": 0.3244, "step": 28300 }, { "epoch": 4.45, "learning_rate": 2.1803921568627452e-06, "loss": 0.2688, "step": 28400 }, { "epoch": 4.47, "learning_rate": 2.1176470588235296e-06, "loss": 0.2043, "step": 28500 }, { "epoch": 4.49, "learning_rate": 2.054901960784314e-06, "loss": 0.3008, "step": 28600 }, { "epoch": 4.5, "learning_rate": 1.9921568627450982e-06, "loss": 0.281, "step": 28700 }, { "epoch": 4.52, "learning_rate": 1.9294117647058825e-06, "loss": 0.32, "step": 28800 }, { "epoch": 4.53, "learning_rate": 1.8666666666666669e-06, "loss": 0.2855, "step": 28900 }, { "epoch": 4.55, "learning_rate": 1.8039215686274512e-06, "loss": 0.2691, "step": 29000 }, { "epoch": 4.56, "learning_rate": 1.7411764705882353e-06, "loss": 0.2638, "step": 29100 }, { "epoch": 4.58, "learning_rate": 1.6784313725490198e-06, "loss": 0.2563, "step": 29200 }, { "epoch": 4.6, "learning_rate": 1.615686274509804e-06, "loss": 0.3033, "step": 29300 }, { "epoch": 4.61, "learning_rate": 1.5529411764705885e-06, "loss": 0.3405, "step": 29400 }, { "epoch": 4.63, "learning_rate": 1.4901960784313726e-06, "loss": 0.2761, "step": 29500 }, { "epoch": 4.64, "learning_rate": 1.427450980392157e-06, "loss": 0.2985, "step": 29600 }, { "epoch": 4.66, "learning_rate": 1.3647058823529413e-06, "loss": 0.2562, "step": 29700 }, { "epoch": 4.67, "learning_rate": 1.3019607843137256e-06, "loss": 0.3165, "step": 29800 }, { "epoch": 4.69, "learning_rate": 1.23921568627451e-06, "loss": 0.3435, "step": 29900 }, { "epoch": 4.71, "learning_rate": 1.1764705882352942e-06, "loss": 0.2907, "step": 30000 }, { "epoch": 4.72, "learning_rate": 1.1137254901960786e-06, "loss": 0.2614, "step": 30100 }, { "epoch": 4.74, "learning_rate": 1.0509803921568629e-06, "loss": 0.2767, "step": 30200 }, { "epoch": 4.75, "learning_rate": 9.88235294117647e-07, "loss": 0.2349, "step": 30300 }, { "epoch": 4.77, "learning_rate": 9.254901960784314e-07, "loss": 0.3014, "step": 30400 }, { "epoch": 4.78, "learning_rate": 8.627450980392157e-07, "loss": 0.284, "step": 30500 }, { "epoch": 4.8, "learning_rate": 8.000000000000001e-07, "loss": 0.2762, "step": 30600 }, { "epoch": 4.82, "learning_rate": 7.372549019607844e-07, "loss": 0.2582, "step": 30700 }, { "epoch": 4.83, "learning_rate": 6.745098039215686e-07, "loss": 0.3047, "step": 30800 }, { "epoch": 4.85, "learning_rate": 6.11764705882353e-07, "loss": 0.3008, "step": 30900 }, { "epoch": 4.86, "learning_rate": 5.490196078431373e-07, "loss": 0.2766, "step": 31000 }, { "epoch": 4.88, "learning_rate": 4.862745098039216e-07, "loss": 0.3692, "step": 31100 }, { "epoch": 4.89, "learning_rate": 4.235294117647059e-07, "loss": 0.2916, "step": 31200 }, { "epoch": 4.91, "learning_rate": 3.6078431372549024e-07, "loss": 0.317, "step": 31300 }, { "epoch": 4.93, "learning_rate": 2.9803921568627456e-07, "loss": 0.267, "step": 31400 }, { "epoch": 4.94, "learning_rate": 2.3529411764705883e-07, "loss": 0.2752, "step": 31500 }, { "epoch": 4.96, "learning_rate": 1.7254901960784313e-07, "loss": 0.2656, "step": 31600 }, { "epoch": 4.97, "learning_rate": 1.0980392156862746e-07, "loss": 0.2457, "step": 31700 }, { "epoch": 4.99, "learning_rate": 4.705882352941176e-08, "loss": 0.2846, "step": 31800 }, { "epoch": 5.0, "eval_accuracy": 0.9948888888888889, "eval_loss": 0.02364550158381462, "eval_runtime": 47.8839, "eval_samples_per_second": 187.955, "eval_steps_per_second": 23.494, "step": 31875 }, { "epoch": 5.0, "step": 31875, "total_flos": 1.976187420555264e+19, "train_loss": 0.3736599187514361, "train_runtime": 3816.438, "train_samples_per_second": 66.816, "train_steps_per_second": 8.352 } ], "max_steps": 31875, "num_train_epochs": 5, "total_flos": 1.976187420555264e+19, "trial_name": null, "trial_params": null }