remove default to ignore safetensors

[felipemello1]

Context

What is the purpose of this PR? Is it to

• add a new feature
• fix a bug
• update tests and/or documentation
• other (please add here)

bad default

Testing:
Downloaded models

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/torchtune/2042

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

• [DomainsOnly] Jobs fail with GLIBC version not found

✅ No Failures

As of commit b05689b with merge base e9fd56a ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[felipemello1]

unfortunately, 8b uses meta ckpt, and 70b uses hg ckpt. Ideally, we should update their configs to all use hf, so we can always ignore consolidated. To avoid adding too much info here, i suggested that the user can just remove the ignore patterns flag, downloading all models.

[joecummings]

Remove the mention of this in kaggle_download as well.

[calvinpelletier]

Not important but just for cleanliness, you could remove any --ignore-patterns None in the configs now that it is the default:

find /path/to/torchtune/recipes/configs -type f -name "*.yaml" -exec sed -i 's/ --ignore-patterns None//g' {} +

[felipemello1]

Not important but just for cleanliness, you could remove any --ignore-patterns None in the configs now that it is the default:

find /path/to/torchtune/recipes/configs -type f -name "*.yaml" -exec sed -i 's/ --ignore-patterns None//g' {} +

yessir