Skip to content

Pull requests: pytorch/rl

New pull request New
79 Open 1,989 Closed
79 Open 1,989 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[Bug] Fixes bug in GAE advantage estimation when gammalmbda is a Tensor bug Something isn't working CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Suitable for minor Suitable to be integrated in minor release (no new feature)
#2773 opened Feb 7, 2025 by louisfaury Loading…
2 of 10 tasks
7
Fix tutos CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2772 opened Feb 7, 2025 by vmoens Loading…
3
[BugFix] NonTensor should not convert anything to numpy CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2771 opened Feb 7, 2025 by vmoens Loading…
3
[BugFix] Avoid calling reset during env init CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2770 opened Feb 7, 2025 by vmoens Loading…
3
[BugFix] Use brackets to get non-tensor data in gym envs CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2769 opened Feb 7, 2025 by vmoens Loading…
3
[BE] Remove deprec specs from tests CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2767 opened Feb 7, 2025 by vmoens Loading…
2
[DRAFT] ppo chess with llm and ConditionalPolicySwitch to sunfish bot CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2763 opened Feb 5, 2025 by mikaylagawarecki • Draft
12
[Feature] TensorDictPrimer with single default_value callable CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. enhancement New feature or request
#2732 opened Jan 30, 2025 by vmoens Loading…
4
[Feature] ConditionalPolicySwitch transform CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. enhancement New feature or request
#2711 opened Jan 21, 2025 by vmoens Loading…
1
[Example] Self-play chess PPO example CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Examples
#2709 opened Jan 21, 2025 by vmoens Loading…
1
[WIP] Compute lp during loss execution CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2688 opened Jan 10, 2025 by vmoens Loading…
3
[CI] Fix conda on windows CI Has to do with CI setup (e.g. wheels & builds, tests...) CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2676 opened Dec 20, 2024 by vmoens Loading…
10 tasks
3
[Tutorial] MCTS CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2673 opened Dec 19, 2024 by vmoens Loading…
3
First draft for modular Hindsight Experience Replay Transform CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. enhancement New feature or request
#2667 opened Dec 19, 2024 by dtsaras • Draft
3 of 10 tasks
31
[Tutorial] Beam search with GPT models CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. tutorials
#2623 opened Dec 2, 2024 by vmoens Loading…
6
[Feature] PPOTrainer CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2550 opened Nov 11, 2024 by vmoens Loading…
3
[Feature] habitat env from config bug Something isn't working CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. enhancement New feature or request
#2539 opened Nov 6, 2024 by vmoens Loading…
10 tasks
3
[CI] Fix windows upload wheels CI Has to do with CI setup (e.g. wheels & builds, tests...) CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2507 opened Oct 21, 2024 by vmoens Loading…
3
[Feature] Gymnasium 1.0 compatibility CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Environments Adds or modifies an environment wrapper
#2473 opened Oct 9, 2024 by vmoens Loading…
26
[Examples] boiler plate code for multi-turn reward for RLHF CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. enhancement New feature or request
#2467 opened Oct 5, 2024 by rghosh08 Loading…
3 of 10 tasks
1
4
[Algorithm] Update scripts with compile CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2449 opened Sep 23, 2024 by vmoens Loading…
1
[Feature] RB compability with compile CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. enhancement New feature or request
#2426 opened Sep 9, 2024 by vmoens Loading…
3
[CI] Add benchmarks to test runs CI Has to do with CI setup (e.g. wheels & builds, tests...) CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2410 opened Sep 2, 2024 by vmoens Loading…
3
[Feature] non-functional SAC loss CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#2393 opened Aug 13, 2024 by vmoens Loading…
3
[Feature] use_vmap=False for SAC CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. enhancement New feature or request
#2392 opened Aug 13, 2024 by vmoens Loading…
3
ProTip! Mix and match filters to narrow down what you’re looking for.