{"id":24487158,"url":"https://github.com/manchery/iql-pytorch","last_synced_at":"2025-07-04T23:41:03.530Z","repository":{"id":60353534,"uuid":"470820524","full_name":"Manchery/iql-pytorch","owner":"Manchery","description":"Unofficial PyTorch implementation (replicating paper results) of Implicit Q-Learning (In-sample Q-Learning) for offline RL","archived":false,"fork":false,"pushed_at":"2024-11-04T15:09:34.000Z","size":267,"stargazers_count":23,"open_issues_count":0,"forks_count":1,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-04-13T21:06:29.918Z","etag":null,"topics":["implicit-q-learning","offline-reinforcement-learning","pytorch","reinforcement-learning"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Manchery.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2022-03-17T02:25:29.000Z","updated_at":"2025-03-01T09:20:52.000Z","dependencies_parsed_at":"2024-11-04T16:33:22.555Z","dependency_job_id":null,"html_url":"https://github.com/Manchery/iql-pytorch","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Manchery/iql-pytorch","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Manchery%2Fiql-pytorch","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Manchery%2Fiql-pytorch/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Manchery%2Fiql-pytorch/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Manchery%2Fiql-pytorch/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Manchery","download_url":"https://codeload.github.com/Manchery/iql-pytorch/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Manchery%2Fiql-pytorch/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":263636513,"owners_count":23492270,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["implicit-q-learning","offline-reinforcement-learning","pytorch","reinforcement-learning"],"created_at":"2025-01-21T15:33:36.884Z","updated_at":"2025-07-04T23:41:03.494Z","avatar_url":"https://github.com/Manchery.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# IQL Implementation in PyTorch\n\n## IQL\n\nThis repo is an unofficial implementation of **Implicit Q-Learning (In-sample Q-Learning)** in PyTorch.\n\n```\n@inproceedings{\n    kostrikov2022offline,\n    title={Offline Reinforcement Learning with Implicit Q-Learning},\n    author={Ilya Kostrikov and Ashvin Nair and Sergey Levine},\n    booktitle={International Conference on Learning Representations},\n    year={2022},\n    url={https://openreview.net/forum?id=68n2s9ZJWF8}\n}\n```\n\n**Note**: Reward standardization (_We standardize MuJoCo locomotion task rewards by dividing by the difference of returns of the best and worst trajectories in each dataset_) used in [official implementation](https://github.com/ikostrikov/implicit_q_learning/blob/09d700248117881a75cb21f0adb95c6c8a694cb2/train_offline.py#L51C18-L51C18) is missed in this implementation. One can easily add it by itself.\n\n## Train\n\n### Gym-MuJoCo\n\n```\npython main_iql.py --env halfcheetah-medium-v2 --expectile 0.7 --temperature 3.0 --eval_freq 5000 --eval_episodes 10 --normalize\n```\n\n### AntMaze\n\n```\npython main_iql.py --env antmaze-medium-play-v2 --expectile 0.9 --temperature 10.0 --eval_freq 50000 --eval_episodes 100\n```\n\n## Results\n\n![mujoco_results](imgs/mujoco_results.png)\n\n![antmaze_results](imgs/antmaze_results.png)\n\n## Acknowledgement\n\nThis repo borrows heavily from [sfujim/TD3_BC](https://github.com/sfujim/TD3_BC) and [ikostrikov/implicit_q_learning](https://github.com/ikostrikov/implicit_q_learning).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmanchery%2Fiql-pytorch","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmanchery%2Fiql-pytorch","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmanchery%2Fiql-pytorch/lists"}