-
Notifications
You must be signed in to change notification settings - Fork 69
Pull requests: meta-pytorch/torchforge
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Enable online writing to WandB in MAST jobs
CLA Signed
This label is managed by the Meta Open Source bot.
#628
opened Dec 8, 2025 by
daniellepintz
Loading…
Fix example command in This label is managed by the Meta Open Source bot.
sandbox/rl_trainer/main.py
CLA Signed
#622
opened Dec 5, 2025 by
daniellepintz
Loading…
[GRPO] Make dataloader deterministic
CLA Signed
This label is managed by the Meta Open Source bot.
#609
opened Dec 1, 2025 by
daniellepintz
Loading…
feat: Reduce reference model memory with with parallel logprob computation
CLA Signed
This label is managed by the Meta Open Source bot.
#608
opened Nov 30, 2025 by
gitlost-murali
Loading…
Update model in This label is managed by the Meta Open Source bot.
tests/sandbox/vllm/qwen2_5_32b.yaml
CLA Signed
#607
opened Nov 27, 2025 by
daniellepintz
•
Draft
[Prototype] Multi-turn GRPO for blackjack with OpenEnv
CLA Signed
This label is managed by the Meta Open Source bot.
#603
opened Nov 20, 2025 by
felipemello1
Loading…
Dp/aws fair
CLA Signed
This label is managed by the Meta Open Source bot.
#598
opened Nov 20, 2025 by
daniellepintz
•
Draft
Refactor and Improve ReinforceLoss implementation
CLA Signed
This label is managed by the Meta Open Source bot.
#583
opened Nov 17, 2025 by
bohdan-nd
Loading…
[WIP][RFC] Multi-turn toolcall
CLA Signed
This label is managed by the Meta Open Source bot.
#567
opened Nov 13, 2025 by
felipemello1
Loading…
Reward Ensemble for RewardActor (Pre-Weaver)
CLA Signed
This label is managed by the Meta Open Source bot.
#566
opened Nov 13, 2025 by
hgKang02
Loading…
[wip][do not review] enable pipeline rl
CLA Signed
This label is managed by the Meta Open Source bot.
Qwen3 Config
CLA Signed
This label is managed by the Meta Open Source bot.
#545
opened Nov 10, 2025 by
pbontrager
•
Draft
Adds integration tests to CI
CLA Signed
This label is managed by the Meta Open Source bot.
#539
opened Nov 7, 2025 by
allenwang28
•
Draft
Add Multi-Node Distributed Training Support for SLURM Clusters
CLA Signed
This label is managed by the Meta Open Source bot.
#528
opened Nov 5, 2025 by
HosseinKaviani-H
Loading…
On Policy Distillation
CLA Signed
This label is managed by the Meta Open Source bot.
#527
opened Nov 5, 2025 by
joecummings
•
Draft
[RFC] - Config is code
CLA Signed
This label is managed by the Meta Open Source bot.
#512
opened Oct 30, 2025 by
felipemello1
Loading…
[wip] Add DeepseekV3 SFT config
CLA Signed
This label is managed by the Meta Open Source bot.
#511
opened Oct 30, 2025 by
daniellepintz
•
Draft
Use smaller runner for docs build
CLA Signed
This label is managed by the Meta Open Source bot.
#470
opened Oct 20, 2025 by
joecummings
Loading…
Install enroot for gpu unit tests
CLA Signed
This label is managed by the Meta Open Source bot.
#456
opened Oct 17, 2025 by
allenwang28
Loading…
Docs Content Part 2: Concepts
CLA Signed
This label is managed by the Meta Open Source bot.
#449
opened Oct 17, 2025 by
AlannaBurke
Loading…
[don't review, debug purpose] Comment out metric logger related statements in grpo.
CLA Signed
This label is managed by the Meta Open Source bot.
NOT_FOR_REVIEW
PR's from Core Maintainers, not intended for review or landing
Previous Next
ProTip!
Add no:assignee to see everything that’s not assigned.