Hackable, with as little abstraction as possible. Done for my own purposes, feel free to rip.
Every model should have its own runnable logic. Seperate. Not shared! Each file does A Thing. The adaptibility of huggingface's code is incredibly bad, due to over abstraction and incidental complexity. Therefore, DIY. Right now, I'm still improving it as I have time.
This is designed to rely on HF for the files, and loading of the files. Otherwise, that is where the dependency ends.
- The code is the tool. Edit it as you see fit!
- All h*ggingface imports will be placed in quarantine. Model pass files will contain no references
- I will not be addressing issues
- Not a single kwargs shall be observed
- One model forward pass = one function call. Simple as.
python ./llama.py