huggingface / accelerate

🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP and DeepSpeed support
https://huggingface.co/docs/accelerate
Apache License 2.0
7.34k stars 875 forks source link

Fix DeepSpeed config validation error by changing `stage3_prefetch_bucket_size` value to an integer #2814

Closed adk9 closed 4 weeks ago

adk9 commented 1 month ago

This PR changes the zero_optimization.stage3_prefetch_bucket_size config value to be an integer. With the impending Pydantic v2 update in DeepSpeed, we run into validation errors if the value is a float. The documentation states that this value should be an integer.

cc @muellerzr

HuggingFaceDocBuilderDev commented 1 month ago

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.