●Change8

Migrating to Accelerate v1.11.0

Version v1.11.0 introduces 1 breaking change. This guide details how to update your code.

Released: 10/20/2025

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:

Accelerator.autocast()FindTiedParametersResultfp8_configmodel.set_requires_gradient_syncget_balanced_memoryget_parameters_from_modulestorch.distributed.barrier

Breaking Changes

●Issue #1

Support for Python 3.9 has been dropped as it reached End-of-Life. Users must upgrade to Python 3.10 or newer.

Migration Steps

  1. 1
    If you are using Python 3.9, you must upgrade your environment to Python 3.10 or newer.
  2. 2
    To use TE MXFP8 support, configure `use_mxfp8_block_scaling` in your `fp8_config`.
  3. 3
    When training on MPS devices, ensure you have torch >= 2.8 for fp16 or torch >= 2.6 for bf16 if using mixed precision.

Release Summary

This release introduces support for TransformerEngine MXFP8 and enables FP16/BF16 training on MPS devices. It also drops support for Python 3.9 and brings numerous stability and feature updates across FSDP and nd-parallelism.

Need More Details?

View the full release notes and all changes for Accelerate v1.11.0.

View Full Changelog