logs9:Fix msec data issue

Fix msec/data issue 2018/04/24

• Log 1: what specific output am I working on right now?
  • msec/data graphs increase on every iteration, there's a chance it's a bug. So we have to investigate.
• Log 2: thinking out loud - e.g. hypotheses about the current problem, what to work on next
  • Run small loop of RL and check.
  • log more what's increasing
  • Observations save() and data iterator was the cause

    • iterator_ops.py:352: UserWarning: An unusually high number of Iterator.get_next() calls was detected. This often indicates that Iterator.get_next() is being called inside a training loop, which will cause gradual slowdown and eventual resource exhaustion. If this is the case, restructure your code to call next_element = iterator.get_next() once outside the loop, and use next_element as the input to some computation that is invoked inside the loop. warnings.warn(GET_NEXT_CALL_WARNING_MESSAGE)

• Log 3: a record of currently ongoing runs along with a short reminder of what question each run is supposed to answer
  • It was actually a really bad issue, I should not have used get_next every iteration. get_next returns get_next op in graph. So run(next_op) actually returns the next dataset. By fixing this. I saw flat msec/data graph.
• Log 4: results of runs (TensorBoard graphs, any other significant observations), separated by type of run (e.g. by the environment the agent is being trained in)