●Change8

Migrating to Transformers v5.10.1

Version v5.10.1 introduces 2 breaking changes. This guide details how to update your code.

Released: 6/3/2026

2
Breaking Changes
1
Migration Steps
12
Affected Symbols

⚠️ Check Your Code

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

Gemma4 vision poolerAudio Language Models (ALMs)Gemma4AltCLIPChineseClipBlip-2WhisperOvis2MoshiTrainer.compute_losscreate_bidirectional_maskBark voice preset embeddings

Breaking Changes

●Issue #1

The Gemma4 vision pooler now casts inputs to float32 before scaling to prevent float16 overflow (inf saturation) with large checkpoints, which may cause minor numerical differences in outputs for users running Gemma-4 vision models in float16.

●Issue #2

Audio Language Models (ALMs) now have a dedicated base model class without a language modeling head, aligning them with the design of Vision Language Models (VLMs); users relying on the previous model class structure should update their code to use the new base model class where appropriate.

Migration Steps

  1. 1
    Users relying on the previous Audio Language Model (ALM) class structure should update their code to use the new dedicated base model class without a language modeling head where appropriate.

Release Summary

This hotfix release addresses issues from the yanked v5.10.0, introduces several new multimodal and specialized models (Gemma4 Unified, Sapiens2, DeepSeek-OCR-2, Mellum), and includes significant fixes for model parallelism and quantization.

Need More Details?

View the full release notes and all changes for Transformers v5.10.1.

View Full Changelog