Skip to content

Pull requests: THUDM/slime

New pull request New
40 Open 1,129 Closed
40 Open 1,129 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: rename samples to seq_lens and remove dead code
#1453 opened Jan 19, 2026 by JackXu0 Loading…
[debug] fixed bug in sft_rollout
#1451 opened Jan 18, 2026 by MikaStars39 Loading…
[Feature] Add multi task logging in slime
#1439 opened Jan 17, 2026 by yitianlian Loading…
Fix multi‑turn loss masking, clarify qwen25/qwen3 types, and strengthen mask tests
#1418 opened Jan 14, 2026 by Daucloud Loading…
issue #1414 fix: correct reward normalization for unequal group sizes
#1415 opened Jan 14, 2026 by ccggddmm Loading…
3
[1/2] feat/add_fault_torlance
#1404 opened Jan 13, 2026 by lilei199908 Loading…
8
support Opencua click dataset
#1393 opened Jan 12, 2026 by kafkayu Loading…
3 of 4 tasks
Add support for per token reward/advantages with a custom_reward_post_process_path
#1389 opened Jan 12, 2026 by vpj Loading…
1
[2/2] add fault torlance
#1311 opened Jan 3, 2026 by lilei199908 Loading…
13
[Feature] Add rollout concurrency argument for full async training
#1310 opened Jan 3, 2026 by yitianlian Loading…
[data][feat] add large dataset support
#1298 opened Dec 31, 2025 by SwordFaith Loading…
12
Handle deepscaler answers without markers
#1226 opened Dec 26, 2025 by cklxx Loading…
1
Fix ruff hook and update pre-commit hooks
#1206 opened Dec 24, 2025 by ParagEkbote Loading…
3
Feat(router): add oai interface support for router
#1203 opened Dec 24, 2025 by ChangyiYang Loading…
10
minor fix test scripts, add default value
#1199 opened Dec 23, 2025 by yueming-yuan • Draft
Integrate Sonic-Moe in FSDP
#1176 opened Dec 22, 2025 by ChangyiYang • Draft
2
[feature] update fp8 weight from megatron fp8 tensor
#1173 opened Dec 21, 2025 by fy1214 • Draft
[FEATURE] Add tool call support for multi-turn SFT with delta-based loss masking
#1159 opened Dec 20, 2025 by Surya-Gunukula Loading…
fix: fix 8B VLM true on policy issue run-ci-short
#1155 opened Dec 19, 2025 by nanjiangwill Loading…
1
[FSDP][1/n] Support LoRA training for FSDP backend.
#1140 opened Dec 17, 2025 by GuanxingLu Loading…
4 tasks
18
[On Policy Distillation] resolve log prob dimension mismatch in on-policy distillation with CP > 1
#1135 opened Dec 17, 2025 by Yuchen-Cao Loading…
5
Optimize CP loss communication
#1102 opened Dec 12, 2025 by cklxx Loading…
4
feat: Support list-of-dicts format for multimodal message content
#1037 opened Dec 5, 2025 by ppraneth Loading…
2
[WIP] Add quantization config -- make DSv3 torch_dist -> hf conversion work
#949 opened Nov 26, 2025 by vwxyzjn • Draft
RDMA Support for the weight transferring from Megatron to SGL
#932 opened Nov 25, 2025 by JensenFire • Draft
1 of 2 tasks
12
ProTip! Follow long discussions with comments:>50.