Bump sglang from 0.4.6.post5 to 0.5.17

[dependabot]

Bumps sglang from 0.4.6.post5 to 0.5.17.

Release notes

Sourced from sglang's releases.

v0.5.17

Highlights

582 PRs from 194 contributors.

Kimi K3 day-0 support: A 2.8T-parameter multimodal LatentMoE (896 experts, top-16, routed in a 3584-dim latent space) with a 1M-token context, 69 KDA linear-attention layers interleaved with 24 MLA layers, and a MoonViT3d vision tower, shipping as a native MXFP4 checkpoint. SGLang serves it from day 0 with DCP, DSpark speculative decoding, chunked-prefill PP with TP decode, KDA-aware prefix caching, HiCache L2 over DCP, LoRA on the quantized weights, and reasoning, tool-call and OpenAI-compatible serving, verified on NVIDIA GB300 and AMD MI35x (#32541, #32828, #32890, #33025, #33112, blog, cookbook, roadmap).

MiniMax-H3 day-0 support: MiniMax's video generation model that produces a video and a synchronized stereo audio track in one request, served natively on SGLang-Diffusion across all three public task profiles: text-to-video-and-audio (t2va), first/last-frame conditioning (fl2va), and image/video/audio reference conditioning (ref2va, which also covers video-to-video). Verified on B200 (TP2 + Ulysses4), H100 (TP2 + Ulysses2), AMD MI300X and MI355X (Ulysses1/2/4/8), and 2x RTX 5090 with layerwise offload (#33275, cookbook).

Other new models added: EmbeddingGemma and LFM2.5 embedding models, nvidia/MiniMax-M3-NVFP4, plus cookbook recipes for Poolside's Laguna-S-2.1 family and Inkling-Small.

Initial support for the Rust frontend: Migrates the front half of the server, everything from network ingress up to the point a tokenized request is handed to the GPU scheduler, from Python to a multi-threaded Rust implementation (#29799).

DCP communication backends and q-replicate (Helix): The DeepSeek-MLA decode context-parallel path gains pluggable comm backends. a2a exchanges packed attention output plus fp32 LSE in a single NCCL collective per layer, with fp8 KV carried as uint8 byte transport; fi_a2a delegates the cross-rank exchange to the FlashInfer MNNVL kernel on GB200. --dcp-replicate-q-proj projects full-head Q locally and skips the per-layer Q head-dim all-gather. Select with --dcp-comm-backend {ag_rs, a2a, fi_a2a} (#21637).

DWDP for MoE prefill: A new prefill parallelism strategy that prefetches peer expert weights over NVLink P2P and computes all experts locally, removing EP all-to-all token dispatch. On 4x B200 with gpt-oss-120b, prefill-only, DWDP4 reaches 1.92x over DEP4 at MNT 32K / ISL 32K, and 506K vs 329K tok/s (1.54x) at saturation (CONC=128, ISL=8K). Enable with --dwdp-size; the authors mark it early-development (#29778).

Session-reference-aware Unified Radix Cache: For agentic and RL-rollout workloads, requests can carry a stable session_id so eviction knows which prefixes an active session still references, instead of evicting purely by cache policy. Release the references with /close_session. Opt in with --enable-session-radix-cache (#29173).

SM90 FP8 MegaMoE for DeepSeek-V4: Adds the DeepGEMM MegaMoE A2A path on SM90 for DeepSeek-V4-Flash/Pro FP8, including the pre-dispatch JIT kernel and FP8 expert weight preparation. Guarded behind SGLANG_OPT_USE_DEEPGEMM_MEGA_MOE=1 (#29016).

Faster large-MoE model loading: Oversized or non-contiguous CPU weight views were driving pathological H2D transfers, with DeepSeek-V4-Pro TP8 spending 27 to 32 minutes in H2D on some ranks. Copying those views into contiguous storage before H2D cuts full model loading from about 35 minutes to 6m20s (5.6x), and GPT-OSS-20B BF16 from 545s to 70s (7.8x), with Qwen3.5-397B measured at 1.93x to 2.3x. Opt in with SGLANG_MOE_COPY_WEIGHT_VIEWS_BEFORE_H2D, off by default (#32315).

Lower DeepSeek-V4 memory on AMD: Removing unnecessary expert padding drops MI355X FP4 MoE model weights from 159.07 GB to 112.36 GB, and bringing the HIP compress-state pool into the memory_saver KV_CACHE region lets colocated RL reclaim it, cutting the measured training-phase footprint from about 143 GiB to 87 GiB per GPU (#31450, #31747).

Faster engine recovery: Large-model restarts cost 3 to 6+ minutes today, about 6.5 minutes for Qwen3-235B FP8 on 4 GPUs, because weights reload from storage and CUDA graphs recapture. A weight-cache daemon holds weights per GPU so a restarting engine can recover from cache instead (#27139).

Lower host overhead in hybrid-linear MTP decode: Under spec-v2 overlap scheduling each decode step runs draft, verify and extend CUDA graphs, and the eager seams between them become GPU idle time at low concurrency. This trims that host work so the host stays off the critical path (#32219).

Dependencies: flashinfer 0.6.15.post1 (#31927), sgl-deep-gemm 0.1.5.post1 (#32345, #33143), helion 1.4 (#32562), mooncake 0.3.12.post1 (#32302), dynamo-tokenizers 1.7.0 (#32981). PyTorch stays at 2.11.0 and the CUDA base image at 13.0.1.

Full release notes by category below; breaking changes and known issues are at the end.

New Model Support

Model	Type	PRs	Cookbook
Kimi K3	Autoregressive (Multimodal)	#32541, #32828, #32890, #33025	link
MiniMax-H3	Diffusion	#33275	link
MiniMax-M3-NVFP4	Autoregressive	#31989
EmbeddingGemma	Autoregressive (Embedding)	#32375, #32383	link
LFM2.5	Autoregressive (Embedding)	#28691	link

Kimi K3

• [Kimi] Support kimi-k3: #32541 ⭐
• [Kimi] Support DCP + DSpark (ported from kimi-k3 branch): #32828
• [Kimi K3] Add reasoning, tool-call, and OpenAI serving support: #33025
• feat(kernels): port standalone Kimi K3 kernels: #32890
• [Feat] DCP + HiCache L2 Support (ported from kimi-k3): #33112

... (truncated)

Commits

• 2948168 [Cherry-pick to release/v0.5.17] fix(PP): size the mamba pool per pipeline st...
• 58be7db [Cherry-pick to release/v0.5.17] [Kimi-K3] Allow DSPARK verify on cutedsl_mla...
• 4c06614 [Cherry-pick to release/v0.5.17] [AMD] Fuse Kimi-K3 attn-residual aggregation...
• 8af8fba [Cherry-pick to release/v0.5.17] Fix MXFP4 scale placeholder initialization (...
• 9e3287e [Cherry-pick to release/v0.5.17] docker: add Kimi K3 artifacts and build hpc-...
• d3cba91 fix(gdn): skip the -1 padding sentinel in the chunked extend kernel (#33810)
• 20cef18 Fix Mistral-Large-3 EAGLE draft skipping DeepseekV2Model.init
• 7163b4a [Cherry-pick to release/v0.5.17] [CI] Temporarily disable prefill cuda graph ...
• f2890ad [Cherry-pick to release/v0.5.17] Fix Nightly NV CI (#33564) (#33779)
• 4fbc160 [Cherry-pick to release/v0.5.17] [DCP] Match the replicated draft KV pool's p...
• Additional commits viewable in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)