●Change8

Migrating to vLLM v0.8.0rc1

Version v0.8.0rc1 introduces 1 breaking change. This guide details how to update your code.

Released: 3/17/2025

1
Breaking Changes
3
Migration Steps
7
Affected Symbols

⚠️ Check Your Code

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

AsyncLLMLLMEngineFlashPagedAttentionbenchmark_serving.pyVLLM_ATTENTION_BACKENDprefix_prefillvllm:cache_config_info

Breaking Changes

●Issue #1

vLLM no longer sets the global seed automatically (#14274). This may lead to non-deterministic results in environments relying on the previous behavior.

Migration Steps

  1. 1
    Manually set the `seed` parameter in your configuration or API calls if you require reproducible results, as the global seed is no longer set by default.
  2. 2
    Update scripts using `benchmark_serving.py` to remove the deprecated `--dataset` flag.
  3. 3
    If using FlashInfer, ensure it is explicitly installed if `VLLM_ATTENTION_BACKEND` is set to avoid startup errors.

Release Summary

This release introduces Expert Parallelism for DeepSeek models, a new /score endpoint for embeddings, and significant V1 engine enhancements including parallel sampling. It also removes global seed setting, requiring users to manually define seeds for reproducibility.

Need More Details?

View the full release notes and all changes for vLLM v0.8.0rc1.

View Full Changelog