v0.2.5-ggml

fix docker run

update to v0.1.53

upgrade to v0.1.53

update for latest ggml-v3

v0.1.52-rc1

update to latest 0.1.52 llama.cpp

re-release update w latest llama-cpp-python release and drop openblas support

v0.1.50-rc3

give release step write permissions

gpu build w cublas

v0.1.49-rc7

don't worry about openblas

build w ubuntu 18.04 to match HF spaces runtime

v0.1.49-rc6

approve apt install

separate cpu and gpu releases for py38

v0.1.49-rc5

fix path for wheels

v0.1.49-rc2 release with permissions to publish the release

allow github actions the permission to write