Why training accuracy is 0.000000?

[alanMachineLeraning]

epoch 0, total_step 300, total loss is 51.74 , inference loss is 37.28, weight deacy loss is 14.46, training accuracy is 0.000000, time 33.227 samples/sec epoch 0, total_step 320, total loss is 44.49 , inference loss is 30.03, weight deacy loss is 14.46, training accuracy is 0.000000, time 33.665 samples/sec epoch 0, total_step 340, total loss is 51.21 , inference loss is 36.76, weight deacy loss is 14.46, training accuracy is 0.000000, time 33.431 samples/sec epoch 0, total_step 360, total loss is 49.44 , inference loss is 34.99, weight deacy loss is 14.45, training accuracy is 0.000000, time 32.708 samples/sec epoch 0, total_step 380, total loss is 46.34 , inference loss is 31.89, weight deacy loss is 14.45, training accuracy is 0.000000, time 32.569 samples/sec epoch 0, total_step 400, total loss is 47.22 , inference loss is 32.77, weight deacy loss is 14.45, training accuracy is 0.062500, time 32.474 samples/sec epoch 0, total_step 420, total loss is 45.56 , inference loss is 31.11, weight deacy loss is 14.45, training accuracy is 0.000000, time 32.665 samples/sec epoch 0, total_step 440, total loss is 46.80 , inference loss is 32.35, weight deacy loss is 14.45, training accuracy is 0.000000, time 33.084 samples/sec

[zh7117]

at begging, it's hardly impossible for network to do correct classification, so the training acc is always 0.

[sosoIrving]

at begging, it's hardly impossible for network to do correct classification, so the training acc is always 0.

May you tell me when the network to do the correct classification?

[vancuonghoang]

did anyone fix this issue ? I'm trying to train in about 200-300 epochs, but accuracy always 0.00000