Repository Issues
sgl-project/sglang
SGLang is a high-performance serving framework for large language models and multimodal models.
Issues
Open
KeyError: 'model.layers.14.mlp.shared_expert.gate_gate_up_proj.weight'
good first issuehelp wanted
Has a beginner-friendly labelContributing guide available
12 comments1 reaction1 assignee
Open
[Feature] Simplify `tree_speculative_sampling_target_only`
good first issue
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
3 comments0 reactions0 assignees
Open
[Feature] Overlap Spec Support
enhancementgood first issuehelp wantedhigh priority
Why recommendedHas a beginner-friendly label · Contributing guide available
Has a beginner-friendly labelContributing guide available
27 comments25 reactions2 assignees
Open
[Feature] Support piggyback server load report.
good first issue
Why recommendedHas a beginner-friendly label · Contributing guide available
Has a beginner-friendly labelContributing guide available
4 comments0 reactions1 assignee
Open
[Feature] Integrate KVPress in order to try more KV cache optimization heuristics easily
good first issue
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
9 comments2 reactions0 assignees
Open
Reducing the time to import sglang
good first issuehigh priority
Why recommendedHas a beginner-friendly label · Contributing guide available
Has a beginner-friendly labelContributing guide available
6 comments4 reactions1 assignee
Open
[Performance] Remove instances of `torch.nonzero()` where appropriate
good first issueperformance
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
9 comments1 reaction0 assignees
Open
[Feature] Create benchmark and dataset usage scripts for embedding models
good first issuehelp wanted
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
6 comments0 reactions0 assignees
Open
[Feature] Implement Image Number Limitation for Image Processors
Multi-modalgood first issue
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
3 comments0 reactions0 assignees
Open
[Tracking] Improve Multimodal CI coverage
Multi-modalcigood first issueperformance
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
3 comments5 reactions0 assignees
Open
[Bug] Mllama ambiguous multi-image behaviour
Multi-modalgood first issuehelp wanted
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
3 comments0 reactions0 assignees
Open
[Feature] Benchmark with audio input
Multi-modalgood first issuehelp wanted
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
4 comments0 reactions0 assignees
Open
[Feature] Enable EPLB in Draft Models
good first issue
Why recommendedHas a beginner-friendly label · Contributing guide available
Has a beginner-friendly labelContributing guide available
6 comments0 reactions1 assignee
Open
[Feature] Enable Two-Batch Overlap in Draft Models
good first issue
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
5 comments0 reactions0 assignees
Open
[Bug] Kimi VL GPU memory usage too high
Multi-modalgood first issue
Why recommendedHas a beginner-friendly label · Contributing guide available
Has a beginner-friendly labelContributing guide available
17 comments0 reactions2 assignees
Open
[Feature] Increase server payload size limit
good first issue
Why recommendedHas a beginner-friendly label · Contributing guide available
Has a beginner-friendly labelContributing guide available
9 comments1 reaction3 assignees
Open
[Feature] Support `abort_request` in the router
good first issue
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
2 comments0 reactions0 assignees
Closed
[Feature] Support `return_hidden_states` in async engine
good first issue
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
1 comment0 reactions0 assignees
Open
[Feature] Support varied input formats for remaining VLM
good first issuehelp wanted
Why recommendedHas a beginner-friendly label · Contributing guide available
Has a beginner-friendly labelContributing guide available
7 comments0 reactions3 assignees
Open
Why recommendedNo assignee yet · Has a beginner-friendly label
No assignee yetHas a beginner-friendly labelContributing guide available
4 comments1 reaction0 assignees