David Foster

Hi - thanks I'll bear that in mind. Chess would be an interesting one to include for sure, though there will certainly be other implementations out there (e.g. [LeelaZero](https://github.com/leela-zero/leela-zero)) that...

I have just released a new `tensorflow_2` branch which maps everything across to using Keras within tensorflow 2.0. Can you see if you still get the error on this branch?

Yes agreed - it's tricky as usually the model input shape is dependent on the number of players (e.g. 4 player sushi go requires the agent to keep track of...

For 1. see the original paper, under the section entitled 'Photo generation from paintings' "For painting→photo, we find that it is helpful to introduce an additional loss to encourage the...

I had the same issue - for me what solved it was ensuring the response is in one-hot encoded form (i.e. for binary problem, the input is [[0,1],[1,0]] instead of...

I've tried including the following lines in the install.sh script (not sure that --system is a valid flag, as mentioned in the official docs?): ``` python3.6 -m pip install --target...

Thanks for the suggestion - where would I do that? I've tried entering this at the top of both MyBot.py and model.py but I'm still getting the same error. ```...

Also seeing this bug - means that `ChatInterface` cannot be used with reload functionality currently. Can't see any workaround, other than not to use auto-reload.

Hi @araffin - thanks for the reply! Does VecEnv parallelise the gradient computation, or just the env part? https://twitter.com/hardmaru/status/1260852988475658242 I've got PPO + MPI working really well on a multicore...

If you're using Docker, you can get rid of this message by including `--cap_add=SYS_PTRACE` in the `docker run`