Fork of LittleLittleCloud/Torchsharp-llama
- Obtain, and download the model weights. Make sure to download the pth version, ie, the one without -hf prefix
- llama-2-7b: https://huggingface.co/meta-llama/Llama-2-7b
- llama-2-7b-chat: https://huggingface.co/meta-llama/Llama-2-7b-chat
-
Pass in the path to the weights directory as a program argument.
# example ${PATH_TO_LLAMA}/llama-2-7b-chat