v3.2.1

Also available at: DockerHub

v3.2.0

Also available at: DockerHub

v3.1.2

Also available at: DockerHub

v3.1.0

Also available at: DockerHub

v2.2.1

Update benchmark.py

v2.0.0

• Upgraded upstream tag enables Llama 3.1 in ollama
• Support for AmpereOne platform
• Breaking change: due to changed weight type IDs it is now required to re-quantize models to Q8R16 and Q4_K_4 formats with current llama-quantize tool.

v1.2.6

Create README.md

v1.2.3

• The rebase is to allow llama-cpp-python to pick up upstream CVE fix (GHSA-56xg-wfcc-g829)
• Experimental support for Q8R16 quantized format with optimized matrix multiplication kernels
• CMake files updated to build llama.aio on AmpereOne

v1.2.2

Release notes:

• Fix llama-3 end of token issue
• Update server to support ollama (v0.1.33)
• llama.aio docker support server mode by default

SHA-256 hashes:

• 6c580006a8faf7b73a424b0020f1bda2684aa7e1796182f68bfa8b7fee08d991 llama_cpp_python-0.2.63-cp311-cp311-linux_aarch64.whl
• 1ffde8093abe18f638fb89273dd56664dd7ff6b8c82383099ea620d18ab562a7 llama_aio_v1.2.2_b769bc1.tar.gz