-
Couldn't load subscription status.
- Fork 3.2k
Pull requests: sgl-project/sglang
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[MoE] Remove logic that prevents the user from selecting MoE backend for specific models
#12291
opened Oct 28, 2025 by
Jonahcb
Loading…
1 of 4 tasks
Support of piecewise graph compilation for prefill on NPU
#12287
opened Oct 28, 2025 by
Vladimir221
Loading…
2 of 4 tasks
Revert "Fix potential eos bug on decode instance when PD is enabled"
run-ci
#12282
opened Oct 28, 2025 by
hnyls2002
Loading…
feat: preview filename from tuning_fused_moe_triton.py
#12276
opened Oct 28, 2025 by
lianakoleva
Loading…
4 tasks done
[Feature] PD-Multiplexing Context and Scheduler, lazy import spatial.
run-ci
#12275
opened Oct 28, 2025 by
ykcombat
Loading…
1 of 4 tasks
Super tiny add unit tests on cutlass moe single-batch overlap
run-ci
#12273
opened Oct 28, 2025 by
fzyzcjy
Loading…
4 tasks
[bugfix] Fix tcp port conflict(zmq.error.ZMQError: Address already in use) when zmq socket is bound to random port
#12272
opened Oct 28, 2025 by
wcsjtu
Loading…
3 tasks
Super tiny fix expert distribution dump error
run-ci
#12271
opened Oct 28, 2025 by
fzyzcjy
Loading…
4 tasks
Super tiny add UT for copy_to_gpu_no_ce
run-ci
#12270
opened Oct 28, 2025 by
fzyzcjy
Loading…
4 tasks
doc: improve modelopt error description
run-ci
#12269
opened Oct 28, 2025 by
lianakoleva
Loading…
4 tasks done
[Bug fix] Fix severe memory waste issue with torch.empty pin_memory
#12266
opened Oct 28, 2025 by
sjtushenhai
Loading…
4 tasks
[BugFix][Qwen2.5-VL]: fix cu_seqlens in qwen2.5-vl
#12261
opened Oct 28, 2025 by
gjghfd
Loading…
4 tasks
[hotfix] missing
w13_weight_fp8 and w2_weight_fp8 in UE8M0 requantization
run-ci
#12259
opened Oct 28, 2025 by
ch-wan
Loading…
4 tasks
fix: Add default value for backend in sample_mmmu_requests
#12256
opened Oct 28, 2025 by
ZailiWang
Loading…
fix(moe): Add global cache reuse to prevent OOM during chunked prefill
run-ci
#12251
opened Oct 28, 2025 by
liusy58
Loading…
4 tasks
[NVIDIA] Add CI workloads for GB200
run-ci
#12242
opened Oct 28, 2025 by
kaixih
Loading…
2 of 4 tasks
Add continuous_usage_stats support for streaming responses
run-ci
#12241
opened Oct 28, 2025 by
BBuf
Loading…
[VLM] Optimize qwen_vl preprocess_video
run-ci
#12240
opened Oct 28, 2025 by
yuan-luo
Loading…
4 tasks
Previous Next
ProTip!
What’s not been updated in a month: updated:<2025-09-28.