Use httplib to download model from an URL when the libcurl is disabled

Note: LLAMA_HTTPLIB is OFF, cannot build llama-server. Hint: to skip building server, set -DLLAMA_BUILD_SERVER=OFF

Author: JamePeng

.github/workflows/build-wheels-cu124-cu126-win-basic.yml

@@ -69,7 +69,7 @@ jobs:
           $env:VERBOSE = '1'
           $env:CMAKE_ARGS = '-DGGML_CUDA=on -DCMAKE_CUDA_ARCHITECTURES=' + $env:CUDAARCHVER + ' -DCMAKE_BUILD_PARALLEL_LEVEL=' + $env:MAX_JOBS
           $env:CMAKE_ARGS = "-DGGML_CUDA_FORCE_MMQ=on -DCUDA_SEPARABLE_COMPILATION=on $env:CMAKE_ARGS"
-          $env:CMAKE_ARGS = "-DENABLE_CCACHE=on -DLLAMA_CURL=off $env:CMAKE_ARGS"
+          $env:CMAKE_ARGS = "-DENABLE_CCACHE=on -DLLAMA_CURL=off -DLLAMA_HTTPLIB=on $env:CMAKE_ARGS"
 
           # Basic options for compiling without AVX instructions
           if ($env:AVXVER -eq 'Basic') {

.github/workflows/build-wheels-cu124-cu126-win.yml

@@ -69,7 +69,7 @@ jobs:
           $env:VERBOSE = '1'
           $env:CMAKE_ARGS = '-DGGML_CUDA=on -DCMAKE_CUDA_ARCHITECTURES=' + $env:CUDAARCHVER + ' -DCMAKE_BUILD_PARALLEL_LEVEL=' + $env:MAX_JOBS
           $env:CMAKE_ARGS = "-DGGML_CUDA_FORCE_MMQ=on -DCUDA_SEPARABLE_COMPILATION=on $env:CMAKE_ARGS"
-          $env:CMAKE_ARGS = "-DENABLE_CCACHE=on -DLLAMA_CURL=off $env:CMAKE_ARGS"
+          $env:CMAKE_ARGS = "-DENABLE_CCACHE=on -DLLAMA_CURL=off -DLLAMA_HTTPLIB=on $env:CMAKE_ARGS"
 
           if ($env:AVXVER -eq 'AVX') {
             $env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DGGML_AVX=on -DGGML_AVX2=off -DGGML_FMA=off -DGGML_F16C=off'

.github/workflows/build-wheels-cu124-linux.yml

@@ -62,8 +62,8 @@ jobs:
 
           # Add project-specific and feature flags
           CMAKE_ARGS="-DGGML_CUDA=on -DCMAKE_CUDA_ARCHITECTURES='70-real;75-real;80-real;86-real;87-real;89-real'"
-          CMAKE_ARGS="-DGGML_CUDA_FORCE_MMQ=ON ${CMAKE_ARGS}"
-          CMAKE_ARGS="${CMAKE_ARGS} -DLLAMA_CURL=ON"
+          CMAKE_ARGS="-DGGML_CUDA_FORCE_MMQ=on ${CMAKE_ARGS}"
+          CMAKE_ARGS="${CMAKE_ARGS} -DLLAMA_CURL=on -DLLAMA_HTTPLIB=on"
 
           if [ "${AVXVER}" = "AVX" ]; then
             CMAKE_ARGS="${CMAKE_ARGS} -DGGML_AVX=on -DGGML_AVX2=off -DGGML_FMA=off -DGGML_F16C=off"

.github/workflows/build-wheels-cu126-linux.yml

@@ -62,8 +62,8 @@ jobs:
 
           # Add project-specific and feature flags
           CMAKE_ARGS="-DGGML_CUDA=on -DCMAKE_CUDA_ARCHITECTURES='70-real;75-real;80-real;86-real;87-real;89-real'"
-          CMAKE_ARGS="-DGGML_CUDA_FORCE_MMQ=ON ${CMAKE_ARGS}"
-          CMAKE_ARGS="${CMAKE_ARGS} -DLLAMA_CURL=ON"
+          CMAKE_ARGS="-DGGML_CUDA_FORCE_MMQ=on ${CMAKE_ARGS}"
+          CMAKE_ARGS="${CMAKE_ARGS} -DLLAMA_CURL=on -DLLAMA_HTTPLIB=on"
 
           if [ "${AVXVER}" = "AVX" ]; then
             CMAKE_ARGS="${CMAKE_ARGS} -DGGML_AVX=on -DGGML_AVX2=off -DGGML_FMA=off -DGGML_F16C=off"

.github/workflows/build-wheels-cu128-linux.yml

@@ -62,8 +62,8 @@ jobs:
 
           # Add project-specific and feature flags
           CMAKE_ARGS="-DGGML_CUDA=on -DCMAKE_CUDA_ARCHITECTURES='75-real;80-real;86-real;87-real;89-real;90-real;100-real;101-real;120-real'"
-          CMAKE_ARGS="-DGGML_CUDA_FORCE_MMQ=ON ${CMAKE_ARGS}"
-          CMAKE_ARGS="${CMAKE_ARGS} -DLLAMA_CURL=ON"
+          CMAKE_ARGS="-DGGML_CUDA_FORCE_MMQ=on ${CMAKE_ARGS}"
+          CMAKE_ARGS="${CMAKE_ARGS} -DLLAMA_CURL=on -DLLAMA_HTTPLIB=on"
 
           if [ "${AVXVER}" = "AVX" ]; then
             CMAKE_ARGS="${CMAKE_ARGS} -DGGML_AVX=on -DGGML_AVX2=off -DGGML_FMA=off -DGGML_F16C=off"

.github/workflows/build-wheels-cu128-win.yml

@@ -69,7 +69,7 @@ jobs:
           $env:VERBOSE = '1'
           $env:CMAKE_ARGS = '-DGGML_CUDA=on -DCMAKE_CUDA_ARCHITECTURES=' + $env:CUDAARCHVER + ' -DCMAKE_BUILD_PARALLEL_LEVEL=' + $env:MAX_JOBS
           $env:CMAKE_ARGS = "-DGGML_CUDA_FORCE_MMQ=on -DCUDA_SEPARABLE_COMPILATION=on $env:CMAKE_ARGS"
-          $env:CMAKE_ARGS = "-DENABLE_CCACHE=on -DLLAMA_CURL=off $env:CMAKE_ARGS"
+          $env:CMAKE_ARGS = "-DENABLE_CCACHE=on -DLLAMA_CURL=off -DLLAMA_HTTPLIB=on $env:CMAKE_ARGS"
 
           if ($env:AVXVER -eq 'AVX') {
             $env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DGGML_AVX=on -DGGML_AVX2=off -DGGML_FMA=off -DGGML_F16C=off'