Trying to work my way through the docker+huggingface tutorial “LLM Everywhere”
I do not find it straightforward to follow at all - is there anywhere an easier tutorial?
The chosen Llama2 model is large, possibly exceeding my MacBook storage - why not choose a smaller model for a tutorial. Not sure if it can even run on a MacBook, would be nice to know system requirements.
Llama2 needs extra registration - so with docker, HF and Llama one is juggling 3 registrations. I get unexplained time-outs in the ‘quick demo’ . Suspecting that a needed Llama2 registration is the cause for that I submitted my Llama2 registration. I got a message that this will now be “reviewed” by Meta. Now what.
Scrolling down hundreds of more lines of python code. In the conclusion it says it is all ‘convenient’.
Again, anywhere a tutorial with more of a knack for teaching?
Would love to build docker images somewhere else than my MacBook - but where?