I have been trying to download Llama 3.1 8B. Failed to download every time. Tried everything. I do not know what is happening. It gets stuck. If I restart it gets stuck to a new point. Any advice would be appreciated.
Hi! Could you share your code so we can have more context?
Do you have access on Huggingface? Some models require approval from their uploaders on HF: meta-llama/Llama-3.1-8B · Hugging Face . You then need to make sure your HF token is set before trying to run something using 8B.
I’d advise to try if you can load it with transformers.AutoModelForCausalLM, if you can’t that’s a hf issue, not a nnsight one
Hi All,
Sorry for the late reply. It took a while to investigate. I think the problem is with the the computer where I was working. I am still working to fix it. I tried Automodelforcausallm but that did not help.
yeah if loading with huggingface doesn’t work, it’s not an nnsight issue