Fix bug for checkpoint saving on multi node training setting (#28078) #6254

	name: Update Transformers metadata

	on:
	push:
	branches:
	- main
	- update_transformers_metadata*

	jobs:
	build_and_package:
	runs-on: ubuntu-22.04
	defaults:
	run:
	shell: bash -l {0}

	steps:
	- uses: actions/checkout@v3

	- name: Setup environment
	run: \|
	pip install --upgrade pip
	pip install datasets pandas==2.0.3
	pip install .[torch,tf,flax]

	- name: Update metadata
	run: \|
	python utils/update_metadata.py --token ${{ secrets.LYSANDRE_HF_TOKEN }} --commit_sha ${{ github.sha }}

Provide feedback