I'm learning TensorFlow and trying to apply it on MNIST database, and surprised to see already in 2nd epoch (e1 in the screenshot) the accuracy on whole test set reaches 95% from just 9.4% during e0 (before any updates to weight & bias!)!
Is it possible such a sharp rise in accuracy?
- code: https://github.com/vibhorj/tf/blob/master/so/mnist-3-relu.py
- batch size: 100
- epochs: 10000
- 5 layer network (4 hidden, 1 output; 200 / 100 / 60 / 30 neirons)
- lr = 0.001
- AdamOptimizer
- relu & softmax actiovation functions
Considering batch size of 100 & mnist training samples # = 55000. there must have been 550 updates before e1.
Is there anything wrong i'm doing ?
Thanks a lot in advance!