search

yhyu13 / AlphaGOZero-python-tensorflow

Congratulation to DeepMind! This is a reengineering implementation (on behalf of many other git repo in /support/) of DeepMind's Oct19th publication: [Mastering the Game of Go without Human Knowledge]. The supervised learning approach is more practical for individuals. (This repository has single purpose of education only)
MIT License
340 stars 113 forks source link

Remove NoneType in results preprocessing #18

Open awilliamson opened 6 years ago

awilliamson commented 6 years ago

None types in results for preprocessing will result in an error when trying to return. Filtering out the None types beforehand appears to fix this, and preprocessing - alongside #17 - continues to actually write now.

awilliamson commented 6 years ago

Don't merge yet. This has knock on effects when training due to how chunking is done. As expected shape is 100000. I believe a more involved approach to filtering out these NoneTypes is required. Will update when possible.

CRuJia commented 4 years ago

When I remove NoneType in results preprocessing, I have another error that the training trunk shape is not 100000. How to resolve this error?