-
Notifications
You must be signed in to change notification settings - Fork 3.6k
Pull requests: sgl-project/sglang
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
feat(GroupCoordinator): Avoid creating an excessive number of invalid process groups.
#14360
opened Dec 3, 2025 by
CLFutureX
Loading…
Fix sgl-router silently parse selector wrongly causing OME fail to discover pods
model-gateway
run-ci
#14359
opened Dec 3, 2025 by
fzyzcjy
Loading…
6 tasks
[DLLM] Add documentation for diffusion LLM
documentation
Improvements or additions to documentation
#14358
opened Dec 3, 2025 by
ClawSeven
Loading…
[Perf] Enable Flashinfer autotune by default
#14357
opened Dec 3, 2025 by
elvischenv
Loading…
6 tasks
[do not merge] Update transformers package version to 5.0.0rc0
dependencies
Pull requests that update a dependency file
run-ci
#14356
opened Dec 3, 2025 by
yhyang201
Loading…
6 tasks
Feature/add vae path to cli doc#14004
diffusion
SGLang Diffusion
documentation
Improvements or additions to documentation
#14355
opened Dec 3, 2025 by
baonudesifeizhai
Loading…
6 tasks
multimodal: precompute hash for MultimodalDataItem
documentation
Improvements or additions to documentation
Multi-modal
multi-modal language model
vlm
#14354
opened Dec 3, 2025 by
sufeng-buaa
Loading…
2 of 6 tasks
feat(dsv32): better error handling for DeepSeek-v3.2 encoder
#14353
opened Dec 3, 2025 by
jimmy-evo
Loading…
1 of 6 tasks
[FIX] trtllm-moe-fp4-renorm for Qwen series models
#14350
opened Dec 3, 2025 by
samuellees
Loading…
6 tasks
add doc for quantized kv cache
documentation
Improvements or additions to documentation
quant
LLM Quantization
#14348
opened Dec 3, 2025 by
b8zhong
Loading…
[Performance] Optimize radix cache eviction performance
run-ci
#14339
opened Dec 3, 2025 by
YiXR
Loading…
3 of 6 tasks
remove unecessary dual stream token threshold from the rest of models (qwen moe, kimi linear, etc.)
run-ci
#14337
opened Dec 3, 2025 by
b8zhong
Loading…
Add rope kernel in sgl-kernel
run-ci
sgl-kernel
#14334
opened Dec 3, 2025 by
Qiaolin-Yu
Loading…
6 tasks
feat: V32 tool call parsing for no-dsml tag
deepseek
#14332
opened Dec 3, 2025 by
Eva20150932-atlascloud
Loading…
5 tasks
[model-gateway] use worker crate in openai router
model-gateway
run-ci
#14330
opened Dec 3, 2025 by
slin1237
Loading…
6 tasks
Move custom_ops under layers; move _custom_ops.py → custom_all_reduce_ops.py
run-ci
#14326
opened Dec 3, 2025 by
merrymercy
Loading…
[DeepseekV3.2][NSA][Indexer] Fix PAGED top-k transform for NSA indexer chunked execution on H200
run-ci
#14325
opened Dec 3, 2025 by
YAMY1234
Loading…
6 tasks
[Generative Score API] Fix on prefill-only scheduler running batch loss track problem
#14320
opened Dec 2, 2025 by
haNa-meister
Loading…
6 tasks done
[6/n] Fix
num_token_non_padded computation in prefill
npu
#14313
opened Dec 2, 2025 by
yuchengz816-bot
•
Draft
6 tasks
[model-gateway] change sgl-router to sgl-model-gateway
deepseek
dependencies
Pull requests that update a dependency file
documentation
Improvements or additions to documentation
model-gateway
run-ci
#14312
opened Dec 2, 2025 by
slin1237
Loading…
6 tasks
[Fix] add block size logic for sm120 smem size
#14311
opened Dec 2, 2025 by
koush
Loading…
2 of 6 tasks
[SMG][DS32][fix] support dsv32, add role developer
model-gateway
run-ci
#14307
opened Dec 2, 2025 by
jimmy-evo
Loading…
feat(router): Add load-aware fallback to cache-aware policy to prevent hotspots
model-gateway
#14305
opened Dec 2, 2025 by
ppraneth
Loading…
6 tasks
[FIX][DS32]openai protocol: support openai message role: developer
#14304
opened Dec 2, 2025 by
jimmy-evo
Loading…
2 tasks done
Previous Next
ProTip!
Exclude everything labeled
bug with -label:bug.