Confusing message when not having enough space on device

[KennethEnevoldsen]

I didn't have enough space on the device to download the model, but the error I got was:

from sentence_transformers import SentenceTransformer
model = SentenceTransformer("emillykkejensen/EmbeddingGemma-Scandi-300m")
# OSError: Can't load the model for 'emillykkejensen/EmbeddingGemma-Scandi-300m'. If you were trying to load it from 'https://huggingface.co/models', make sure you don't have a local directory with the same name. Otherwise, make sure 'emillykkejensen/EmbeddingGemma-Scandi-300m' is the correct path to a directory containing a file named pytorch_model.bin, tf_model.h5, model.ckpt or flax_model.msgpack.

Which makes it seem like there is a model implementation error rather than a local one.

Removing large files resolves the issue.

[tomaarsen]

Hello!

This error originates from transformers. My guess is that it has overly broad try-except sections that assume that any model loading error might be caused by various different errors, rather than catching specific issues like out-of-space errors and throwing relevant exceptions.
There's not much I can do about it currently as this all happens in transformers. It might also be a bit of a mess to reproduce/test nicely.

• Tom Aarsen

[KennethEnevoldsen]

Ahh totally that makes sense - I will leave this open in case you want to keep it for tracking, but feel free to close it. If I get the issue again I will open an issue in transformers