●Change8

Migrating to vLLM v0.8.0

Version v0.8.0 introduces 3 breaking changes. This guide details how to update your code.

Released: 3/18/2025

3
Breaking Changes
5
Migration Steps
11
Affected Symbols

⚠️ Check Your Code

If you use any of these symbols, you need to read this guide:

VLLM_USE_V1forward_contextgeneration_configSupportsV0OnlyFlashMLALMCacheSGMVBGMVvllm:time_in_queue_requestsvllm:model_forward_time_millisecondsvllm:model_execute_time_milliseconds

Breaking Changes

●Issue #1

The default value of 'seed' is now None. To ensure reproducibility, you must now explicitly set the seed value.

●Issue #2

The 'kv_cache' and 'attn_metadata' arguments have been removed from the model's forward method. Access these via 'forward_context' instead.

●Issue #3

vLLM now defaults 'generation_config' from the model for chat templates and sampling parameters (e.g., temperature), which may change output behavior if not specified.

Migration Steps

  1. 1
    Update code to explicitly set 'seed' if reproducibility is required.
  2. 2
    Refactor custom model forward methods to use 'forward_context' instead of 'kv_cache' and 'attn_metadata' arguments.
  3. 3
    To use Gemma 3, install transformers from the main branch: 'pip install git+https://github.com/huggingface/transformers.git'.
  4. 4
    If V1 engine causes issues, disable it by setting the environment variable 'VLLM_USE_V1=0'.
  5. 5
    Update monitoring systems to replace deprecated vllm metrics.

Release Summary

v0.8.0 enables the V1 engine by default, introduces support for NVIDIA Blackwell and Gemma 3, and significantly optimizes DeepSeek model performance through FlashMLA and Expert Parallelism.

Need More Details?

View the full release notes and all changes for vLLM v0.8.0.

View Full Changelog