DeNA / HandyRL

HandyRL is a handy and simple framework based on Python and PyTorch for distributed reinforcement learning that is applicable to your own environments.

DeNA/HandyRL Issues

Replay Buffer
Updated a year ago2
Large scale training
Closed a year ago3
TypeError: list indices must be integers or slices, not str
Closed 2 years ago2
Multiple Ctrl+C is needed to finish every process and thread in the early phase
Closed 2 years ago3
How to make model consider immediate reward ?
Closed 2 years ago4
num_parallel affecting learning results
Updated 2 years ago1
[REQUESTING OVERVIEW OF DISTRIBUTED HANDYRL]
Updated 3 years ago4
The cpu memory keeps increasing, and when it is full, an error got!
Updated 3 years ago8
Suggestion: AMP support
Updated 3 years ago3
Customize learning rate
Closed 3 years ago3
Kaggle GPU's
Updated 3 years ago2
BrokenPipeError: [Errno 32] Broken pipe
Updated 3 years ago7
Make it a library
Updated 3 years ago3
Going further into deep MARL with halite and HandyRL
Closed 3 years ago5
OSError when running on Windows
Closed 3 years ago4
How to avoid `OSError: [Errno 24] Too many open files`
Closed 3 years ago3
train stop at 690epoch
Closed 3 years ago6
How do i know when ive reached an optimum while training
Updated 3 years ago3
[Question] [Documentation]
Updated 3 years ago
Examples of config.yaml for supported games
Closed 3 years ago8
How is multi-agent handled ?
Closed 3 years ago8
Unexpected results inference time
Closed 3 years ago
HungryGeese environment, too many open files in Ubuntu 18.04
Closed 3 years ago4