Feat update env and add how to run on gpu docs

README.md

@@ -1,4 +1,5 @@
 # Proactive Interventions in Autonomous Reinforcement Learning
+
 [![License](https://img.shields.io/badge/license-MIT-blue.svg)](https://github.com/rlworkgroup/metaworld/blob/master/LICENSE)
 
 This is the official PyTorch implementation of our NeurIPS 2022 paper ["When to Ask for Help: Proactive Interventions in Autonomous Reinforcement Learning"](https://arxiv.org/abs/2210.10765) by [Annie Xie*](https://anxie.github.io/), [Fahim Tajwar*](https://tajwarfahim.github.io/), [Archit Sharma*](https://architsharma97.github.io/), [Chelsea Finn](https://ai.stanford.edu/~cbfinn/). Please see the [project website](https://sites.google.com/view/proactive-interventions) for example results. For any questions/concerns related to the codebase, please reach out to [Fahim Tajwar](mailto:tajwarfahim932@gmail.com).
@@ -18,6 +19,7 @@ If you use this repo in your research, please consider citing our paper:
 ```
 
 ## Installation
+
 Install [MuJoCo](http://www.mujoco.org/) if it is not already the case:
 
 * Obtain a license on the [MuJoCo website](https://www.roboti.us/license.html).
@@ -56,17 +58,20 @@ conda env create -f paint/conda_env.yml
 conda activate paint
 ```
 
-
 ## Running experiments
 
 First, make sure you are in the "paint" directory.
 
 **Train an episodic PAINT agent (maze):**
 
+To run this on cpu config, just run the following command:
+
 ```
 bash run_scripts/maze.sh
 ```
 
+If you wanted to run it via gpu, edit `cfgs/oracle.yaml` and set `device: cuda` and then run the previous command.
+
 **Train a PAINT agent in a continuing task setting (cheetah):**
 
 ```

earl_benchmark/env.yml

@@ -7,7 +7,7 @@ dependencies:
   - pip>=21.0
   - conda>=4.6.7
   - pip:
-    - git+https://github.com/rlworkgroup/metaworld.git@master#egg=metaworld
+    - git+https://github.com/Farama-Foundation/Metaworld.git@04be337a12305e393c0caf0cbf5ec7755c7c8feb
     - mujoco-py==2.0.2.9
     - numpy==1.21.5
     - matplotlib

paint/conda_env.yml

@@ -28,7 +28,7 @@ dependencies:
     - sklearn==0.0
     - matplotlib==3.4.2
     - opencv-python==4.5.3.56
-    - git+https://github.com/rlworkgroup/metaworld.git@master#egg=metaworld
+    - git+https://github.com/Farama-Foundation/Metaworld.git@04be337a12305e393c0caf0cbf5ec7755c7c8feb
     - dm_env
     - bsuite
     - mujoco==2.2.2