Merge pull request #46 from chamalgomes/feat/expand-cpu-instructions

feat: expand CPU instructions

Author: chamalgomes

.github/workflows/llama-build-cuda.yaml

@@ -21,8 +21,8 @@ jobs:
         cuda_config:
           - ver: 13.1.1
             short: cu131
-            arch: 75;80
-        pyver: ["3.13", "3.14"]
+            arch: 75
+        pyver: ["3.14"]
 
     defaults:
       run:
@@ -69,9 +69,9 @@ jobs:
             -DLLAVA_BUILD=off
             -DCMAKE_CUDA_ARCHITECTURES=${{ matrix.cuda_config.arch }}
             -DGGML_CUDA_FORCE_MMQ=OFF
-            -DGGML_AVX2=off
-            -DGGML_FMA=off
-            -DGGML_F16C=off
+            -DGGML_AVX2=on
+            -DGGML_FMA=on
+            -DGGML_F16C=on
             -DLLAMA_BUILD_EXAMPLES=OFF
             -DLLAMA_BUILD_TESTS=OFF
             -DLLAMA_BUILD_SERVER=OFF