This repository has been archived by the owner on Sep 30, 2023. It is now read-only.

WIP: implement locking of model per request #218

Job	Run time
windows-latest-cmake (avx2)	7m 45s
windows-latest-cmake (avx, -DGGML_AVX2=OFF)	7m 52s
windows-latest-cmake (avx512, -DGGML_AVX512=ON)	6m 23s
windows-latest-cmake (openblas, -DGGML_OPENBLAS=ON -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas...	5m 22s
build-macOS	11m 18s
build-ubuntu (avx2, -DGGML_STATIC=On)	2m 44s
build-ubuntu (avx, -DGGML_AVX2=OFF -DGGML_STATIC=On)	2m 33s
build-ubuntu (avx512, -DGGML_AVX512=ON -DGGML_STATIC=On)	3m 42s
build-ubuntu (avx2-openblas, -DGGML_OPENBLAS=ON -DGGML_STATIC=On)	2m 41s
build-ubuntu (avx2-cuda, -DGGML_CUBLAS=ON -DCMAKE_CUDA_COMPILER=/usr/local/cuda/bin/nvcc -DGGML_...	3m 48s
	54m 8s

Provide feedback