Skip to main content

A library desgined for Offline Preference-Based RL algorithms.

Project description

🚧 This repo is subject to major API changes 🚧

WiseRL provides unofficial and banchmarked PyTorch implementations for Offline Preference-Based RL algorithms, including:

  • Oracle-IQL & Oracle AWAC
  • Supervised Finetuning (SFT)
  • BT Model + IQL/AWAC (BT-IQL/AWAC)
  • Contrastive Prefereing Learning (CPL)
  • Inverse Preference Learning + IQL/AWAC (IPL-IQL/AWAC)

Installation

  • clone this repo and install the dependencies
    git clone git@github.com:typoverflow/WiseRL
    cd WiseRL && pip install -e .
    
  • install environment or dataset dependencies
    • for D4RL experiments:
      git clone https://github.com/Farama-Foundation/d4rl.git
      cd d4rl
      pip install -e .
      
    • for metaworld experiments:
      git clone git@github.com:Farama-Foundation/Metaworld
      cd Metaworld && git checkout 04be337a
      pip install -e .
      

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

wiserl-0.0.2.tar.gz (41.4 kB view details)

Uploaded Source

File details

Details for the file wiserl-0.0.2.tar.gz.

File metadata

  • Download URL: wiserl-0.0.2.tar.gz
  • Upload date:
  • Size: 41.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.0.0 CPython/3.12.2

File hashes

Hashes for wiserl-0.0.2.tar.gz
Algorithm Hash digest
SHA256 35f79e7c45af4da3e10bc38f11792eba64da292b069ea5d8b4d4a3201e5f679e
MD5 e1fbbefe35f8374f9a91222fcc460ed0
BLAKE2b-256 39403c0455b015722bf90f01057495562be09c76cd4c2eb8d04b1d69503f1ea5

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page