Skip to content

Pull requests: OpenPipe/ART

New pull request New
71 Open 687 Closed
71 Open 687 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

Synchronize caller PyTorch randomness across TrainerRank replicas
#918 opened Sep 16, 2026 by bradhilton Collaborator • Draft
1
fix: resume SFT dataset iteration at the exact batch
#908 opened Sep 16, 2026 by RaycarlLei Loading…
Account for grouped checkpoint, head, and enclosing forward buffers
#898 opened Sep 15, 2026 by bradhilton Collaborator • Draft
40
Use Forge as the default Serverless Training endpoint
#897 opened Sep 14, 2026 by Kovbo Collaborator • Draft
Retain managed trainer runtime metadata in the GPU image
#892 opened Sep 12, 2026 by bradhilton Collaborator • Draft
1
Share budgeted CUDA cache recovery with gradient handoff
#888 opened Sep 11, 2026 by bradhilton Collaborator • Draft
12
Preserve captured conditioning when retaining sampled lineage
#885 opened Sep 10, 2026 by bradhilton Collaborator • Draft
1
Export Gemma 4 MoE LoRA as per-expert projections
#865 opened Sep 8, 2026 by Kovbo Collaborator • Draft
SkyPilot example and docs
#815 opened Aug 25, 2026 by alex000kim Loading…
Add Qwen 3.8 27B Megatron support
#810 opened Aug 19, 2026 by Kovbo Collaborator • Draft
[codex] Account for OpenAI Responses cached tokens
#795 opened Aug 14, 2026 by Tanisha-Katara Loading…
Integrate topology-portable TrainerRank checkpoints
#791 opened Aug 12, 2026 by bradhilton Collaborator Loading…
docs: document OrcaRouter as a hosted inference option
#787 opened Aug 11, 2026 by XiaoHuo888-hue Loading…
fix: retry transient vLLM adapter reload failures
#778 opened Jul 27, 2026 by yaodong-shen Loading…
1
1
Add exact token content attribution
#769 opened Jul 22, 2026 by bradhilton Collaborator • Draft
2
feat(examples): add skill-rl for training models on markdown instructions
#753 opened Jul 13, 2026 by menonpg Loading…
[codex] migrate GPU stack to CUDA 13
#729 opened Jun 16, 2026 by bradhilton Collaborator • Draft
Add multi-gpu inference
#727 opened Jun 12, 2026 by Kovbo Collaborator Loading…
2
Remove deprecated old_benchmarking utilities
#712 opened Jun 4, 2026 by Kovbo Collaborator Loading…
Add Trackio trace logging
#687 opened May 21, 2026 by abidlabs Loading…
1
Add verifiers integration bridge
#680 opened May 14, 2026 by kkudumu Loading…
8
Add PipelineTrainer checkpoint artifact uploads
#671 opened May 7, 2026 by arcticfly Collaborator Loading…
Fix LocalBackend LoRA registry sync
#662 opened Apr 27, 2026 by bradhilton Collaborator • Draft
1
3
Add LoRA fork weight loading (pre-transformers-v5 base)
#654 opened Apr 16, 2026 by arcticfly Collaborator Loading…
2 of 3 tasks
Fix LocalBackend fork_checkpoint to overwrite initial LoRA for vLLM
#652 opened Apr 13, 2026 by arcticfly Collaborator Loading…
1
ProTip! Type g i on any issue or pull request to go back to the issue listing page.