Instructions to use maderix/llama-65b-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use maderix/llama-65b-4bit with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("maderix/llama-65b-4bit", dtype="auto") - Notebooks
- Google Colab
- Kaggle
Perhaps there was an issue with the 13B model.
#1
by minlik - opened
The size of 13B model should be approximately twice that of the 7B model, but they are the same in the hub.
so maybe it's a fake?
Yeah you're right..seems like a miss from my end. Will reupload the 13b model
maderix changed discussion status to closed