sglang-diffusion-performance

Name: sglang-diffusion-performance
Availability: InStock
Author: annealing-inversion

Benchmark diffusion configurations across SGLang and Diffusers backends.

Updated Jun 4, 2026

Search Tags:#performance #backend #benchmark #gpu #diffusion #sglang

Ready to run

npx skills add https://github.com/annealing-inversion/sglang-kimi-deferral --skill sglang-diffusion-performance-annealing-inversion

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: sglang-diffusion-performance
Source: https://github.com/annealing-inversion/sglang-kimi-deferral/tree/main/python/sglang/multimodal_gen/.claude/skills/sglang-diffusion-performance
Command: npx skills add https://github.com/annealing-inversion/sglang-kimi-deferral --skill sglang-diffusion-performance-annealing-inversion

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Identify the fastest and most VRAM-efficient diffusion configurations for a given model and GPU setup.

Core Features & Use Cases

Comprehensive benchmarking of native SGLang vs Diffusers backends
Quick comparison of parallelism schemes, attention backends, and caching options
Real-world guidance for selecting configurations under VRAM budgets

Quick Start

Run the diffusion performance skill to benchmark and compare native and diffusers backends on your GPU setup.

Frequently Asked Questions about sglang-diffusion-performance

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I find the fastest diffusion configuration for my GPU?▼

To find the fastest diffusion configuration for your GPU, benchmark native SGLang and Diffusers backends. Comparing parallelism schemes, attention backends, and caching decisions identifies the most VRAM-efficient setup for your specific model and hardware.

Can I benchmark SGLang vs Diffusers backends for diffusion models?▼

Yes, you can benchmark SGLang vs Diffusers backends directly to compare performance. The skill analyzes native SGLang pipelines against Diffusers backends across single- and multi-GPU scenarios to determine the fastest generation flags.

What do I need to benchmark diffusion performance with SGLang?▼

You need a CUDA-capable environment, access to SGLang generate with either --backend sglang or --backend diffusers, and the ability to produce performance dumps. This allows the skill to analyze and identify optimal VRAM configurations.

How do I optimize diffusion VRAM usage across multiple GPUs?▼

Optimize diffusion VRAM usage across multiple GPUs by comparing different parallelism schemes. The skill benchmarks multi-GPU configurations against attention backends and caching decisions to find the most efficient setup within your VRAM budget.

Does SGLang support different attention backends for diffusion models?▼

Yes, SGLang supports different attention backends for diffusion models. You can benchmark these attention backends alongside caching decisions and parallelism schemes to identify the fastest and most VRAM-efficient configuration for your setup.