bug fix in HookToCallback

CHANGES.md

@@ -9,7 +9,7 @@ Here are a list of things that were changed, starting from an early version.
 TensorFlow itself also changes API and those are not listed here.
 
 + 2017/03/16. `session_config` option in `TrainConfig` and `PredictConfig` is deprecated.
-	Use `session_creator` to define how to create session instead.
+	Use `session_creator` to define how to create session instead. See [commit](https://github.com/ppwwyyxx/tensorpack/commit/ccae46f4a3ca89dc3df901a338eef8447d19a730).
 + 2017/02/20. The interface of step callbacks are changed to be the same as `tf.train.SessionRunHook`.
 	If you haven't written any custom step callbacks, there is nothing to do. Otherwise please refer
 	to the [existing callbacks](https://github.com/ppwwyyxx/tensorpack/blob/master/tensorpack/callbacks/steps.py).

tensorpack/callbacks/hooks.py

@@ -55,4 +55,4 @@ class HookToCallback(Callback):
         self._hook.after_run(ctx, run_values)
 
     def _after_train(self):
-        self._hook.end()
+        self._hook.end(self.trainer.sess)