Skip to content

Pull requests: pytorch/rl

New pull request New
94 Open 2,641 Closed
94 Open 2,641 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
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[Test] Add tests and benchmarks for collector throughput optimizations Benchmarks rl/benchmark changes CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Collectors Tests Incomplete or broken unit tests
#3567 opened Mar 24, 2026 by vmoens Loading…
2
[Performance] Enable torch.compile(fullgraph=True) for step_and_maybe_reset CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Performance Performance issue or suggestion for improvement
#3566 opened Mar 23, 2026 by vmoens Loading…
2
[Performance] Add fast path for step() and TransformedEnv._step() when _trust_step_output is set CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Performance Performance issue or suggestion for improvement Transforms
#3565 opened Mar 23, 2026 by vmoens Loading…
2
[Performance] Streamline collector inner loop carrier update CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Collectors Performance Performance issue or suggestion for improvement
#3564 opened Mar 23, 2026 by vmoens Loading…
1
[Performance] Add update_traj_ids flag to Collector to skip trajectory tracking CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Collectors Performance Performance issue or suggestion for improvement
#3563 opened Mar 23, 2026 by vmoens Loading…
3
[Performance] Add _trust_step_output flag to skip step validation overhead CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Performance Performance issue or suggestion for improvement Transforms
#3562 opened Mar 23, 2026 by vmoens Loading…
3
[Performance] Add out= parameter to _StepMDP for output buffer reuse CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Performance Performance issue or suggestion for improvement
#3561 opened Mar 23, 2026 by vmoens Loading…
3
[Performance] Add _skip_maybe_reset flag to bypass auto-reset in step_and_maybe_reset CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Performance Performance issue or suggestion for improvement
#3560 opened Mar 23, 2026 by vmoens Loading…
3
Fix vLLM >= 0.17 compatibility: migrate to native WeightTransferConfig API CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. llm/ LLM-related PR, triggers LLM CI tests Modules sota-implementations/ WeightUpdate
#3556 opened Mar 21, 2026 by vmoens Loading…
5
Export torchrl_logger from torchrl.__init__ for GRPO scripts CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#3555 opened Mar 21, 2026 by vmoens Loading…
4
[CI] Add torch_geometric integration tests 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. Environments Adds or modifies an environment wrapper
#3552 opened Mar 14, 2026 by vmoens Loading…
1 of 2 tasks
2
[LLM] Wire MATH and Countdown into GRPO and Expert Iteration scripts CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. sota-implementations/
#3546 opened Mar 5, 2026 by vmoens Loading…
6
[LLM] Add Countdown numbers-game environment CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Documentation Improvements or additions to documentation llm/ LLM-related PR, triggers LLM CI tests
#3545 opened Mar 5, 2026 by vmoens Loading…
6
[LLM] Add MATH (competition mathematics) environment CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Documentation Improvements or additions to documentation llm/ LLM-related PR, triggers LLM CI tests
#3544 opened Mar 5, 2026 by vmoens Loading…
6
[LLM] Simplify IFEval reward aggregator CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. llm/ LLM-related PR, triggers LLM CI tests
#3543 opened Mar 5, 2026 by vmoens Loading…
6
[LLM] Rewrite GSM8K reward function to follow standard GRPO conventions CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. llm/ LLM-related PR, triggers LLM CI tests sota-implementations/
#3542 opened Mar 5, 2026 by vmoens Loading…
6
[CI] Install torchcodec from source 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. Record
#3541 opened Mar 4, 2026 by vmoens Loading…
1
[Feature] PILCO 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. Documentation Improvements or additions to documentation Feature New feature Modules Objectives sota-implementations/ Transforms
#3537 opened Feb 27, 2026 by PSXBRosa Loading…
3 of 6 tasks
1
26
[Feature] Added Genesis Environments to TorchRL 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. Environments Adds or modifies an environment wrapper Feature New feature
#3536 opened Feb 26, 2026 by ParamThakkar123 Loading…
6 of 10 tasks
1
[Feature] Added Lazy implementation of priority updates for replaybuffer prototype CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Feature New feature ReplayBuffers
#3507 opened Feb 13, 2026 by ParamThakkar123 Loading…
3 of 10 tasks
1
1
[Feature] Added support for TDMPC2 dataset 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. Data Data-related PR, will launch data-related jobs Documentation Improvements or additions to documentation Environments Adds or modifies an environment wrapper Feature New feature
#3501 opened Feb 12, 2026 by ParamThakkar123 Loading…
6 of 10 tasks
2
[Feature] Added OpenEnv environments 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. Documentation Improvements or additions to documentation Environments Adds or modifies an environment wrapper Feature New feature llm/ LLM-related PR, triggers LLM CI tests Trainers
#3470 opened Feb 9, 2026 by ParamThakkar123 Loading…
6 of 10 tasks
18
[Feature] Extended Support delayed spec initialization for exploration modules CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Feature New feature Modules
#3450 opened Feb 5, 2026 by ParamThakkar123 Loading…
3 of 10 tasks
1
1
[Feature] Added MCTSPolicyBase, MCTSPolicy, AlphaGoPolicy, AlphaStarPolicy, and MuZeroPolicy CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Documentation Improvements or additions to documentation Feature New feature Modules
#3449 opened Feb 5, 2026 by ParamThakkar123 Loading…
6 of 10 tasks
14
[Algorithm] DPO CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Documentation Improvements or additions to documentation llm/ LLM-related PR, triggers LLM CI tests Objectives
#3427 opened Jan 31, 2026 by vmoens Loading…
5
ProTip! Adding no:label will show everything without a label.