Skip to content

Pull requests: Gen-Verse/OpenClaw-RL

New pull request New
20 Open 32 Closed
20 Open 32 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

fix: skip DDP and optimizer for prm_teacher (inference-only mode)
#115 opened May 14, 2026 by HeJW0303 Loading…
3 tasks done
Add Ascend NPU support
#103 opened Apr 25, 2026 by quancs Loading…
feat: support resuming existing W&B runs
#102 opened Apr 24, 2026 by Zhuang-Zhuang-Liu Loading…
feat: implement single-GPU (24GB) INT4 QLoRA + offload-rollout + external PRM for OpenClaw-Combine and test
#97 opened Apr 20, 2026 by wuwanchun Loading…
fix: repad rollout samples after sample filtering
#79 opened Apr 15, 2026 by xuyangc03 Contributor Loading…
openclaw-tinker: support Qwen3.5 tool calling
#57 opened Mar 31, 2026 by ojus1 Loading…
1
Fix "strip_thinking function" and "default model name for OpenClaw requests"
#53 opened Mar 30, 2026 by zlh-source Loading…
fix: support terminal-bench original-tasks layout
#50 opened Mar 28, 2026 by ningzimu Loading…
fix: correct Tinker API URL in README
#40 opened Mar 19, 2026 by lghupan Loading…
1 task done
1
Add initial Qwen3.5-4B support and SGLang text-runtime patch
#39 opened Mar 19, 2026 by ShenAC-SAC Loading…
feat: upgrade MiniMax default model to M3
#37 opened Mar 18, 2026 by octo-patch Loading…
docs: add Chinese documentation for core RL methods
#33 opened Mar 17, 2026 by guozhihao-224 Loading…
feat(openclaw-rl): add reward model feedback training
#28 opened Mar 14, 2026 by chengzidl Loading…
3
feat: Support single gpu
#27 opened Mar 13, 2026 by wyqsss Loading…
2
fix: add cd to SLIME_ROOT and derive MEGATRON_ROOT dynamically
#21 opened Mar 12, 2026 by JasonOA888 Loading…
2 tasks done
feat: add openclaw-training-free — zero-GPU experience loop for LLMs
#6 opened Mar 5, 2026 by zcyynl Loading…
fix: add cd to SLIME_ROOT before train_async.py in all training scripts
#5 opened Mar 4, 2026 by kaysonyu Loading…
fix: derive Megatron-LM PYTHONPATH from script location
#3 opened Mar 1, 2026 by koriyoshi2041 Loading…
4 tasks done
docs: fix kl-loss-coef default value in instructions
#2 opened Mar 1, 2026 by koriyoshi2041 Loading…
2 tasks done
fix: correct PYTHONBUFFERED typo to PYTHONUNBUFFERED
#1 opened Mar 1, 2026 by koriyoshi2041 Loading…
3 tasks done
ProTip! What’s not been updated in a month: updated:<2026-05-12.