# vLLM v0.2.0 — Release & Upgrade Intelligence
- **Product:** vLLM (https://vllm.ai)
- **Version:** v0.2.0
- **Released:** Sep 28, 2023
- **Impact Score:** 5/10
- **Breaking Changes:** Not verified
- **Canonical Analysis:** https://changelogs.us/vllm/updates/v0.2.0/

## Executive Summary (TL;DR)
vLLM v0.2.0. Source highlight: Up to 60% performance improvement by optimizing de-tokenization and sampler

## New Features
- add option to shorten prompt print in log
- Add Model Revision Support in https://github.com/vllm-project/vllm/pull/1014
- Add pandas to requirements.txt
- added support for quantize on LLM module
- Add safetensors support for quantized models
- Add minimum capability requirement for AWQ

## Improvements
- Up to 60% performance improvement by optimizing de-tokenization and sampler
- Initial support for AWQ (performance not optimized)

## Bug Fixes
- Fix typo in README.md
- [FIX] Minor bug fixes
- Fix detokenization leaving special tokens
- Fix warning message on LLaMA FastTokenizer
- Fix hanging when prompt exceeds limit
- [FIX] Don't initialize parameter by default

## Upgrade Command
```bash
python -m pip install 'vllm==0.2.0'
```