-
Notifications
You must be signed in to change notification settings - Fork 16.6k
Pull requests: deepseek-ai/DeepSeek-V3
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
fix: add repetition penalty to mitigate multi-turn repetition (fixes #1125)
#1129
opened Mar 1, 2026 by
modimihir07
Loading…
fix: use OrderedDict for proper LRU cache eviction in fp8_cast_bf16.py
#1128
opened Mar 1, 2026 by
modimihir07
Loading…
Add Multi-Token Prediction (MTP) support for speculative decoding
#1122
opened Feb 24, 2026 by
dcol91863
Loading…
fix: fix triton kernel tiling and fp8_gemm swizzle
#1098
opened Jan 29, 2026 by
JackeyLove1
Loading…
feat(inference): Add streaming support imports for high-performance L…
#1080
opened Jan 13, 2026 by
sanjay-aravindh
Loading…
Optimize FP8 Triton kernels for better performance
#1066
opened Dec 25, 2025 by
yurekami
Loading…
1 of 3 tasks
Update: Revise SGLang Multi-Token Prediction details link
#1027
opened Oct 29, 2025 by
LucaLow
Loading…
feat: implement UE8M0 scale format support for FP8 inference
#1023
opened Oct 26, 2025 by
Libres-coder
Loading…
Fix: Prevent infinite “A5A5A5...” repetition loop in generate() (Issue #1008)
#1020
opened Oct 21, 2025 by
Ceaser1717
Loading…
Add GitHub Action to auto-handle low-quality or spam issues
#1011
opened Oct 12, 2025 by
ctkqiang
Loading…
Create Open World Car + Ghost Mode Game -Complete Game Design Document
#988
opened Sep 15, 2025 by
rao118417-hue
Loading…
Previous Next
ProTip!
Type g i on any issue or pull request to go back to the issue listing page.