Welcome! Type "help" for available commands.

Qwen3.8-27B - SGLang Documentation | William Callahan - Bookmarks

~/bookmarks

/**/

Qwen3.8-27B - SGLang Documentation

docs.sglang.ioSaved August 17, 202611 min

Technical Documentation

Summary

This documentation provides technical instructions for deploying the Qwen3.8-27B vision-language model using the SGLang framework. It covers hardware compatibility, quantization options, and specialized memory management configurations for hybrid GDN architectures.

Highlights

Supports deployment on high-end hardware including H200, RTX 5090, and DGX Spark
Offers multiple quantization precisions: BF16, FP8, and NVFP4 W4A4
Includes a Mamba SSM memory ratio calculator to optimize GDN state pool sizing
Supports speculative decoding via EAGLE or DSPARK to improve performance

auto-generated

Preview of Qwen3.8-27B - SGLang Documentation

via SGLang Documentation

Context

Audience

Machine Learning Engineers and AI Infrastructure Developers

DomainMachine Learning

FormatTechnical Documentation

AccessFree online

Topics

FP8 Quantization Qwen 3 LLM Deployment SGLang Mamba Architecture

Visit Site All Bookmarks

SGLangQwen 3Mamba ArchitectureSpeculative DecodingFP8 Quantization

~/bookmarks

Qwen3.8-27B - SGLang Documentation

Summary

Highlights

Context

Topics

Related

Qwen3.8-27B - SGLang Documentation

Summary

Highlights

Context

Topics

Related

~/bookmarks

Qwen3.8-27B - SGLang Documentation

Summary

Highlights

Context

Topics

Related

Discover Similar Content

GitHub - bilikaz/qwen38-flash-next-recipe: qwen38-flash-next recipe for solo GB10

GitHub - MiaAI-Lab/Qwen3.8-27B-DGX-Spark-RTX-6000: Qwen3.8 27B for DGX Spark / RTX 6000 PRO

GitHub - julianmb/q38rocm: Qwen 3.8 27B ROCmFP4 on AMD Strix Halo (Ryzen AI Max+ 395). Up to 36 tok/s via MTP Speculation, TurboQuant & Mesa RADV Wave64.

Discover Similar Content

Qwen3.8-27B - SGLang Documentation

Summary

Highlights

Context

Topics

Related

Discover Similar Content

GitHub - bilikaz/qwen38-flash-next-recipe: qwen38-flash-next recipe for solo GB10

GitHub - MiaAI-Lab/Qwen3.8-27B-DGX-Spark-RTX-6000: Qwen3.8 27B for DGX Spark / RTX 6000 PRO

GitHub - julianmb/q38rocm: Qwen 3.8 27B ROCmFP4 on AMD Strix Halo (Ryzen AI Max+ 395). Up to 36 tok/s via MTP Speculation, TurboQuant & Mesa RADV Wave64.