Skip to main content
Back to top
Ctrl
+
K
Getting Started
Quickstart
Installation
Tutorials
Single NPU (Qwen3 8B)
Single NPU (Qwen2.5-VL 7B)
Multi-NPU (QwQ 32B)
Multi-NPU (deepseek-v2-lite-w8a8)
Multi-Node (DeepSeek)
FAQs
User Guide
Feature Support
Supported Models
Environment Variables
Release note
Developer Guide
Contributing
Versioning policy
Evaluation
Using lm-eval
Using OpenCompass
Using EvalScope
Performance Benchmark
User Story
vLLM Ascend User Stories
xxx project uses Ascend vLLM, gain 200% performance enhancement of inference.
Index
