never get the result about binoculars HOT 1 CLOSED

This is because Binoculars runs two 7B sized language models which could take forever on CPUs. I'm sorry but running LMs on the CPU is not recommended and this behavior is expected. I would recommend you use resources like Kaggle or Colab Notebook to get GPU access if you're constrained on resources.

You might wanna check out this HuggingFace tutorial to optimize LM inference on a CPU. Please feel free to send in PR for proper CPU support if you're able to optimize for it.

Closing this issue since not much I can do here. Thanks for your interest in the project. :)

from binoculars.