Coder Social home page Coder Social logo

muzero-pytorch's People

Contributors

bampt avatar dependabot[bot] avatar grifball avatar koulanurag avatar leovandriel avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

muzero-pytorch's Issues

Problems with Ray version

I noticed that in the requirements.txt file ray 1.13.0 is given. But the code is not compatible with this version, notable during training here:

ray.wait(workers, len(workers))

wait() only accepts one positional parameter and even if I use ray.wait(workers, num_returns=len(workers)), I get the following error:

wait() expected a list of ray.ObjectRef, got list containing <class 'ray.actor.ActorHandle'>

Still searching for a solution, but if somebody has a good idea, that would be awesome.

Runtime Error: Lost reference to actor

[2020-05-09 22:42:48,652][root][ERROR][main.py>] ==> Lost reference to actor
Traceback (most recent call last):
File "main.py", line 80, in
train(muzero_config, summary_writer)
File "/home/noa/cloned/muzero-pytorch/core/train.py", line 343, in train
for rank in range(0, config.num_actors)]
File "/home/noa/cloned/muzero-pytorch/core/train.py", line 343, in
for rank in range(0, config.num_actors)]
File "/home/noa/anaconda3/envs/torch/lib/python3.7/site-packages/ray/actor.py", line 103, in remote
return self._remote(args, kwargs)
File "/home/noa/anaconda3/envs/torch/lib/python3.7/site-packages/ray/actor.py", line 123, in _remote
return invocation(args, kwargs)
File "/home/noa/anaconda3/envs/torch/lib/python3.7/site-packages/ray/actor.py", line 112, in invocation
raise RuntimeError("Lost reference to actor")
RuntimeError: Lost reference to actor

config.atari doesn't exist

in main.py, there is if args.case == 'atari': from config.atari import muzero_config,
but there's no atari.py in config folder, just classic_control.py. is this file missing?

About Reward and Value Support

Hi, I am just reading the original Muzero paper and find one detail that keeps me confused. Can I discuss it here?
In the paper, authors said:

We use a discrete support set of size 601 with one support for every integer between โˆ’300 and 300.

I wonder why they choose this support set? Does it have to do with the specific game environment? If I got another game environment other than Atari, (e.g, a game with a maximum reward of 10 at the end of the game). how should I choose the support set?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.