vLLM releases: v0.25.0

TL;DR

vLLM released version 0.25.0 with 558 commits from 232 contributors, including 64 new ones. This update makes Model Runner V2 the default for all dense models, expanding on previous quantized-model support.

Detailed Summary

The vLLM team has released version 0.25.0, which includes 558 commits from 232 contributors, 64 of whom are new. This update makes Model Runner V2 (MRv2) the default for all dense models, enhancing execution paths particularly with quantized-model support. The broader impact is an improved standardization and efficiency in model handling across the vLLM ecosystem.

Key Points

• • Model Runner V2 is now the default for all dense models.
• • This release includes 558 commits from 232 contributors.
• • The new version supports quantized-models.