Text Generation
Transformers
English
AWQ
Quantization

Possible mistake in from_pretrained() usage in example β€” invalid repo_id error

#5
by a6341556 - opened

Hi, I was following the example usage provided for the model mit-han-lab/nunchaku-t5, and I believe there might be an error in the provided from_pretrained() call.

The code:
text_encoder_2 = NunchakuT5EncoderModel.from_pretrained("mit-han-lab/nunchaku-t5/awq-int4-flux.1-t5xxl.safetensors")

throws this error:
huggingface_hub.errors.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name': 'mit-han-lab/nunchaku-t5/awq-int4-flux.1-t5xxl.safetensors'. Use repo_type argument if needed.

Can you please confirm whether the example is incorrect, and what the correct usage should be?

Thanks!

MIT HAN Lab org

Hi, could you confirm which nunchaku version you used? It should be >=v0.3.0.

Lmxyy changed discussion status to closed

Got it β€” thanks for pointing that out! I’ve updated the Nunchaku version and it works now.

Sign up or log in to comment