Ch 16. Using native python list instead of deque for storing experience, increases sampling performance considerably. #204
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
First of all, thanks for an amazing book and even more awesome github repo. It really help me pickup Deep RL. I tried training Breakout-v0 using your code and I found that the training slowed considerably as time went by. Replay buffer size was set to 1 million. Since collections.deque take O(n) time for random access, I believe it's not suitable for random sampling. I wrote a simple replay buffer that plays well with the rest of the code and uses native python list for storing samples. The performance gain is considerably large. I hope you find this change useful.