Enable GGML_BACKEND_DL

Author: gregwhiteley-intelexvision

.github/workflows/release.yml

@@ -87,7 +87,7 @@ jobs:
             -DCMAKE_EXE_LINKER_FLAGS=-Wl,--allow-shlib-undefined \
             -DLLAMA_FATAL_WARNINGS=ON \
             -DGGML_NATIVE=OFF \
-            -DGGML_BACKEND_DL=OFF \
+            -DGGML_BACKEND_DL=ON \
             -DGGML_CUDA_FA_ALL_QUANTS=ON \
             -DGGML_CUDA=ON
           cmake --build build
@@ -159,7 +159,8 @@ jobs:
           cmake -S . -B build -G "Ninja Multi-Config" ^
             -DCMAKE_CUDA_ARCHITECTURES=${{ matrix.architectures }} ^
             -DGGML_NATIVE=OFF ^
-            -DGGML_BACKEND_DL=OFF ^
+            -DGGML_BACKEND_DL=ON ^
+            -DGGML_CPU_ALL_VARIANTS=ON ^
             -DGGML_CUDA_FA_ALL_QUANTS=ON ^
             -DGGML_CUDA=ON ^
             -DCURL_LIBRARY="%CURL_PATH%/lib/libcurl.dll.a" -DCURL_INCLUDE_DIR="%CURL_PATH%/include" ^