GoingMyWay

docker build failed

1

I run the docker to build the image with `nvidia-docker build -t cule_release .` but failed, it seems install NVDIA/apex failed. OS: Ubuntu18 CUDA 11.0 Outputs ``` Sending build context...

Issue on Windows 10 CMD and Poweshell

5

``` File "D:\x\run.py", line 164, in run_sequential bpar = tqdm.tqdm(total=args.t_max) File "C:\Users\x\anaconda3\envs\py37\lib\site-packages\tqdm\std.py", line 1086, in __init__ self.refresh(lock_args=self.lock_args) File "C:\Users\x\anaconda3\envs\py37\lib\site-packages\tqdm\std.py", line 1380, in refresh self.display() File "C:\Users\x\anaconda3\envs\py37\lib\site-packages\tqdm\std.py", line 1514, in display...

I got an issue. ``` File "C:\Users\WWW\AppData\Local\Continuum\anaconda3\lib\site-packages\pdfrw\pdfreader.py", line 594, in __init__ repr(fdata[-20:])) pdfrw.errors.PdfParseError: EOF mark not found: 'È¥ì wÒ\x80ÜyîA\x96Ò\x81ôT\x90µ\x05\x80' ``` when ``` for inpfn in [file_path_bak, sup_file_path]: writer.addpages(PdfReader(inpfn).pages) ``` I...

Sacred makes pytorch training slow down

4

I have been using sacred for a while and I found there is an issue. That is, while training with Pytorch on GPUs, the training speed slows down. OS: Ubuntu;...

ENH: Add a generic plot_predict function

8

## ENV - Python 2.7 - np.**version**: 1.11.0 - pd.**version**: 0.18.0 - scipy.**version**: 0.17.0 - matplotlib.**version**: 1.3.1 - statsmodels.**version**: 0.5.0 ## WHERE I tried the [ARMA code](http://statsmodels.sourceforge.net/devel/examples/notebooks/generated/tsa_arma_0.html) and I got...

[Feature Request] Melting Pot Integration

1

## Motivation Melting Pot: https://github.com/deepmind/meltingpot is a great MARL evaluation tool for MARL research. ## Solution Integrate Melting Pot into Env Pool ## Additional context Melting Pot was built on...

Is the simulation speed slow?

7

I am testing the `WallJump` scenario with the following code ```python import tqdm import mlagents from mlagents_envs.registry import default_registry env_id = "WallJump" env = default_registry[env_id].make() env.reset() # We will only...

One section in the docs is missing

1

I found one section in the docs is missing. There is no `modifying-the-environment-from-python` found. https://github.com/Unity-Technologies/ml-agents/blob/main/docs/Python-API.md#modifying-the-environment-from-python

How can I get rid of this error without updating TF

3

The CUDA in my machine is old and cannot run the newest TF. I got the error ``` ImportError: This version of Gin-Config requires TensorFlow version >= 1.12.0; Detected an...

Dear author, After reading the code and the paper, I am confused about why Imagination MDP is introduced and why imagination horizon is needed. For example, with a trained world...