Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Behind a Firewall: how to download models? #904

Closed
vlavorini opened this issue Feb 6, 2025 · 3 comments
Closed

Behind a Firewall: how to download models? #904

vlavorini opened this issue Feb 6, 2025 · 3 comments
Labels
question Further information is requested

Comments

@vlavorini
Copy link

Question

I am behind a firewall that blocks the Huggingface website, so I cannot download the relative models.

How can I overcome this issue? An alternative way of having the models?

Thank you

@vlavorini vlavorini added the question Further information is requested label Feb 6, 2025
@dolfim-ibm
Copy link
Contributor

We are just about to release a CLI which allows to download all the artifacts needed at runtime. See #876.

The easiest way would be to download the artifacts from an open network and then transfer the folder in your restricted environment.

@simonschoe
Copy link

@dolfim-ibm I am on docling 2.20.0 but still have to specify pipeline_options.artifacts_path = ".../docling-models". From the docs I inferred that the route should have been deprecated and pointing to one folder above docling_models should be sufficient (e.g., models with subfolders docling-models, CodeFormula, etc). Am I missing something here?

@dolfim-ibm
Copy link
Contributor

Your understanding is correct. The folder of artifacts_path should look something like this:

$ ls -1 ~/.cache/docling/models
EasyOcr
HuggingFaceTB--SmolVLM-256M-Instruct
ds4sd--CodeFormula
ds4sd--DocumentFigureClassifier
ds4sd--docling-models

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
question Further information is requested
Projects
None yet
Development

No branches or pull requests

3 participants