llama_cpp v0.1.1

Chore

• Remove debug binary from Cargo.toml

New Features

• add LlamaModel::load_from_file_async

Bug Fixes

• require llama_context is accessed from behind a mutex
  This solves a race condition when several get_completions threads are spawned at the same time
• start_completing should not be invoked on a per-iteration basis
  There's still some UB that can be triggered due to llama.cpp's threading model, which needs patching up.

Commit Statistics

• 5 commits contributed to the release.
• 13 days passed between releases.
• 4 commits were understood as conventional.
• 0 issues like '(#ID)' were seen in commit messages

Commit Details

view details

• Uncategorized
  • Add LlamaModel::load_from_file_async (3bada65)
  • Remove debug binary from Cargo.toml (3eddbab)
  • Require llama_context is accessed from behind a mutex (b676baa)
  • start_completing should not be invoked on a per-iteration basis (4eb0bc9)
  • Update to llama.cpp 0a7c980 (94d7385)