-
Notifications
You must be signed in to change notification settings - Fork 7.6k
Pull requests: sgl-project/sglang
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
kernel: port CUTLASS fp8_scaled_mm to JIT and M-bucket the SM120 tiles
jit-kernel
sgl-kernel
#33216
opened Aug 1, 2026 by
b8zhong
Collaborator
Loading…
2 tasks done
Dump requests sample fraction
documentation
Improvements or additions to documentation
#33215
opened Aug 1, 2026 by
MohithR17
Loading…
4 of 5 tasks
Fix DeepSeek-OCR batching crash on variable local-crop counts
deepseek
#33214
opened Aug 1, 2026 by
pavansivaram
Contributor
Loading…
2 tasks done
feat(openai): support drop_thinking for DeepSeek chat encoding
#33212
opened Aug 1, 2026 by
lkm2835
Contributor
Loading…
1 of 5 tasks
feat: rust sglang server Vertex APIs
dependencies
Pull requests that update a dependency file
#33211
opened Aug 1, 2026 by
mrain
Loading…
5 tasks
feat(scheduler): add lpm anti-starvation aging (#31954)
#33210
opened Aug 1, 2026 by
Claisenn
Loading…
Reduce the padding overhead of unaligned MXFP8 GEMM shapes
quant
LLM Quantization
#33209
opened Aug 1, 2026 by
b8zhong
Collaborator
Loading…
2 tasks done
[MXFP8] Use FlashInfer CUTLASS for dense GEMM on SM120, delete Triton path
jit-kernel
#33208
opened Aug 1, 2026 by
b8zhong
Collaborator
Loading…
fix(sampler): abort requests with fully-NaN logits instead of sampling garbage (#33187)
#33206
opened Aug 1, 2026 by
Claisenn
Loading…
[Kernel] Unify BaseFusedOp and MultiPlatformOp dispatch
documentation
Improvements or additions to documentation
jit-kernel
piecewise-cuda-graph
quant
LLM Quantization
#33205
opened Aug 1, 2026 by
BBuf
Collaborator
Loading…
spec: DSPARK PD-disaggregation via native draft-KV transfer
#33204
opened Aug 1, 2026 by
bobcao3
Loading…
[docs] deterministic inference: add a coverage map, document the numerics env vars, refresh the backend list
documentation
Improvements or additions to documentation
#33203
opened Aug 1, 2026 by
jwoongkim11
Loading…
BUG:skip when spech4 deepseekv4
deepseek
#33202
opened Aug 1, 2026 by
ved1beta
Loading…
3 of 5 tasks
fix(openai): preserve highest-ranked candidate on top_logprobs text c…
#33201
opened Aug 1, 2026 by
Claisenn
Loading…
5 tasks
[Fix] Align prefill CUDA graph token buckets with attention TP size
#33200
opened Aug 1, 2026 by
lkm2835
Contributor
Loading…
1 of 5 tasks
[Bugfix] Fix causal sliding-window attention in MiMo audio tokenizer
Multi-modal
multi-modal language model
#33198
opened Aug 1, 2026 by
CJack812
Loading…
5 tasks done
[Bugfix] Pad DSA topk_indices_offset under prefill CP
#33197
opened Aug 1, 2026 by
FrankMinions
Contributor
Loading…
5 tasks
[dLLM] Schedule dLLM prefill and decode rows into one mixed round
npu
#33196
opened Aug 1, 2026 by
wenxuewuhd
Contributor
Loading…
3 of 5 tasks
[AMD] Fix JIT compile failure in sgl_kernel/warp.cuh
jit-kernel
#33195
opened Aug 1, 2026 by
At1a8
Contributor
Loading…
3 of 5 tasks
[Bench] Fix peak output throughput counting chunks instead of tokens
#33193
opened Aug 1, 2026 by
decajoin
Contributor
Loading…
3 of 5 tasks
Remove redundant call to get_available_gpu_memory
#33191
opened Aug 1, 2026 by
coolljt0725
Loading…
5 tasks
[dLLM] Run several denoise forwards per scheduled FDFO round
#33190
opened Aug 1, 2026 by
wenxuewuhd
Contributor
Loading…
4 of 5 tasks
[Fix] MiMo detector streaming: flush trailing text, survive split bot token
#33189
opened Aug 1, 2026 by
tancheng33
Loading…
[Test] Add unit tests for poll_based_barrier
#33188
opened Aug 1, 2026 by
Steven0706
Loading…
5 tasks done
Previous Next
ProTip!
Type g i on any issue or pull request to go back to the issue listing page.