Update llama.cpp usage to the current CLI

#2
by kashif HF Staff - opened
Hugging Face Biology Research org

Update the llama.cpp examples to the current llama.cpp CLI:

  • Use llama-completion for raw DNA prompts.
  • Speculative serving uses --spec-draft-n-max / --spec-draft-n-min.
  • Clarify that llama-perplexity scores the file as one stream in context-sized chunks.
  • Point to the Python score() for per-sequence log-likelihoods.
kashif changed pull request status to merged

Sign up or log in to comment