Description

Model download hangs after acquiring the first couple GBs. Can hf_xet be the reason?

Modules

pip list | grep  "xet\|transf\|torch"
hf-xet                  1.1.0
pytorch-lightning       2.4.0
pytorch-metric-learning 2.6.0
pytorch-wpe             0.0.1
torch                   2.7.0+cpu
torch-audiomentations   0.11.1
torch-complex           0.4.4
torch_pitch_shift       1.2.5
torchaudio              2.7.0+cpu
torchmetrics            1.5.1
torchutil               0.0.14
torchvision             0.22.0+cpu
transformers            4.51.3

Reproduce

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText
import os
os.environ["HF_TOKEN"] = "xxx"

processor = AutoProcessor.from_pretrained("meta-llama/Llama-4-Scout-17B-16E" )
model = AutoModelForImageTextToText.from_pretrained("meta-llama/Llama-4-Scout-17B-16E" )

(possible)Error log