GitLab

'elipson' should be epsilon

Fix num_residual_units for w28-10

Fixes #1014

* Update resnet to run with tf r0.12 API.
1. tf.image.per_image_whitening -> tf.image.per_image_standardization
2. Use tf.summary to replace tf.image_summary, tf.scalar_summary, tf.merge_all_summaries.

* remove log

"resisual" => "residual"

The moving average object was shared between training and evaluation. As
a result, the cost during evaluation was averaged with the training
cost. This affects only objective, not precision.

Uses more NumPy now :)