-
Couldn't load subscription status.
- Fork 19
logs6:Make RL model work with small model
Higepon Taro Minowa edited this page Apr 18, 2018
·
1 revision
- Log 1: what specific output am I working on right now?
- Make it possible RL model run with small models, so that we can save time.
- Log 2: thinking out loud - e.g. hypotheses about the current problem, what to work on next
- first we should extract parameters for the training loop, so that we can run it with hparams
- make it run with mode == Test
- Log 3: record of currently ongoing runs along with a short reminder of what question each run is supposed to answer
- should commit the change first
- Should have graphs here and think about next steps.
-
-
-
- Log 4: results of runs (TensorBoard graphs, any other significant observations), separated by type of run (e.g. by environment the agent is being trained in)