A Dockerized build for CodeLlama, using llama.cpp and ttyd web terminal

• Drag your models into the folder to be mounted (in my case, CodeLlama-7b)
• Build and run the image, mounting your model (use build.sh)
• Access interactive codeLlama in your web browser on localhost:7681

• Prompts folder is copied into the image