tests : download models only when running ctest (#18843)

Signed-off-by: Adrien Gallouët <angt@huggingface.co>

cmake/common.cmake

@@ -33,25 +33,3 @@ function(llama_add_compile_flags)
         endif()
     endif()
 endfunction()
-
-function(llama_download_model NAME HASH)
-    set(DEST "${CMAKE_BINARY_DIR}/${NAME}")
-    get_filename_component(DEST_DIR "${DEST}" DIRECTORY)
-    file(MAKE_DIRECTORY "${DEST_DIR}")
-    if(NOT EXISTS "${DEST}")
-        message(STATUS "Downloading ${NAME} from ggml-org/models...")
-    endif()
-    file(DOWNLOAD
-        "https://huggingface.co/ggml-org/models/resolve/main/${NAME}?download=true"
-        "${DEST}"
-        TLS_VERIFY ON
-        EXPECTED_HASH ${HASH}
-        STATUS status
-    )
-    list(GET status 0 code)
-    if(NOT code EQUAL 0)
-        list(GET status 1 msg)
-        message(FATAL_ERROR "Failed to download ${NAME}: ${msg}")
-    endif()
-    set(LLAMA_DOWNLOAD_MODEL "${DEST}" PARENT_SCOPE)
-endfunction()

cmake/download-models.cmake

@@ -0,0 +1,21 @@
+get_filename_component(DEST_DIR "${DEST}" DIRECTORY)
+file(MAKE_DIRECTORY "${DEST_DIR}")
+
+if(NOT EXISTS "${DEST}")
+    message(STATUS "Downloading ${NAME} from ggml-org/models...")
+endif()
+
+file(DOWNLOAD
+    "https://huggingface.co/ggml-org/models/resolve/main/${NAME}?download=true"
+    "${DEST}"
+    TLS_VERIFY ON
+    EXPECTED_HASH ${HASH}
+    STATUS status
+)
+
+list(GET status 0 code)
+
+if(NOT code EQUAL 0)
+    list(GET status 1 msg)
+    message(FATAL_ERROR "Failed to download ${NAME}: ${msg}")
+endif()

examples/eval-callback/CMakeLists.txt

@@ -4,10 +4,23 @@ install(TARGETS ${TARGET} RUNTIME)
 target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT})
 target_compile_features(${TARGET} PRIVATE cxx_std_17)
 
-set(TEST_TARGET test-eval-callback)
-if(NOT ${CMAKE_SYSTEM_PROCESSOR} MATCHES "s390x")
-    llama_download_model("tinyllamas/stories15M-q4_0.gguf" SHA256=66967fbece6dbe97886593fdbb73589584927e29119ec31f08090732d1861739)
-else()
-    llama_download_model("tinyllamas/stories15M-be.Q4_0.gguf" SHA256=9aec857937849d976f30397e97eb1cabb53eb9dcb1ce4611ba8247fb5f44c65d)
+if(LLAMA_BUILD_TESTS)
+    if(NOT ${CMAKE_SYSTEM_PROCESSOR} MATCHES "s390x")
+        set(MODEL_NAME "tinyllamas/stories15M-q4_0.gguf")
+        set(MODEL_HASH "SHA256=66967fbece6dbe97886593fdbb73589584927e29119ec31f08090732d1861739")
+    else()
+        set(MODEL_NAME "tinyllamas/stories15M-be.Q4_0.gguf")
+        set(MODEL_HASH "SHA256=9aec857937849d976f30397e97eb1cabb53eb9dcb1ce4611ba8247fb5f44c65d")
+    endif()
+    set(MODEL_DEST "${CMAKE_BINARY_DIR}/${MODEL_NAME}")
+    set(TEST_TARGET test-eval-callback)
+    add_test(NAME ${TEST_TARGET}-download-model COMMAND ${CMAKE_COMMAND}
+        -DDEST=${MODEL_DEST}
+        -DNAME=${MODEL_NAME}
+        -DHASH=${MODEL_HASH}
+        -P ${CMAKE_SOURCE_DIR}/cmake/download-models.cmake
+    )
+    set_tests_properties(${TEST_TARGET}-download-model PROPERTIES FIXTURES_SETUP ${TEST_TARGET}-download-model)
+    add_test(NAME ${TEST_TARGET} COMMAND llama-eval-callback -m "${MODEL_DEST}" --prompt hello --seed 42 -ngl 0)
+    set_tests_properties(${TEST_TARGET} PROPERTIES FIXTURES_REQUIRED ${TEST_TARGET}-download-model)
 endif()
-add_test(NAME ${TEST_TARGET} COMMAND llama-eval-callback -m "${LLAMA_DOWNLOAD_MODEL}" --prompt hello --seed 42 -ngl 0)

tests/CMakeLists.txt

@@ -202,11 +202,24 @@ llama_build_and_test(
 llama_build_and_test(test-regex-partial.cpp)
 
 if (NOT ${CMAKE_SYSTEM_PROCESSOR} MATCHES "s390x")
-    llama_download_model("tinyllamas/stories15M-q4_0.gguf" SHA256=66967fbece6dbe97886593fdbb73589584927e29119ec31f08090732d1861739)
+    set(MODEL_NAME "tinyllamas/stories15M-q4_0.gguf")
+    set(MODEL_HASH "SHA256=66967fbece6dbe97886593fdbb73589584927e29119ec31f08090732d1861739")
 else()
-    llama_download_model("tinyllamas/stories15M-be.Q4_0.gguf" SHA256=9aec857937849d976f30397e97eb1cabb53eb9dcb1ce4611ba8247fb5f44c65d)
+    set(MODEL_NAME "tinyllamas/stories15M-be.Q4_0.gguf")
+    set(MODEL_HASH "SHA256=9aec857937849d976f30397e97eb1cabb53eb9dcb1ce4611ba8247fb5f44c65d")
 endif()
-llama_build_and_test(test-thread-safety.cpp ARGS -m "${LLAMA_DOWNLOAD_MODEL}" -ngl 99 -p "The meaning of life is" -n 128 -c 256 -ub 32 -np 4 -t 2)
+set(MODEL_DEST "${CMAKE_BINARY_DIR}/${MODEL_NAME}")
+
+add_test(NAME test-download-model COMMAND ${CMAKE_COMMAND}
+    -DDEST=${MODEL_DEST}
+    -DNAME=${MODEL_NAME}
+    -DHASH=${MODEL_HASH}
+    -P ${CMAKE_SOURCE_DIR}/cmake/download-models.cmake
+)
+set_tests_properties(test-download-model PROPERTIES FIXTURES_SETUP test-download-model)
+
+llama_build_and_test(test-thread-safety.cpp ARGS -m "${MODEL_DEST}" -ngl 99 -p "The meaning of life is" -n 128 -c 256 -ub 32 -np 4 -t 2)
+set_tests_properties(test-thread-safety PROPERTIES FIXTURES_REQUIRED test-download-model)
 
 llama_build_and_test(test-arg-parser.cpp)
 
@@ -223,7 +236,8 @@ llama_build_and_test(test-backend-sampler.cpp   LABEL "model")
 
 # Test for state restore with fragmented KV cache
 # Requires a model, uses same args pattern as test-thread-safety
-llama_build_and_test(test-state-restore-fragmented.cpp LABEL "model" ARGS -m "${LLAMA_DOWNLOAD_MODEL}")
+llama_build_and_test(test-state-restore-fragmented.cpp LABEL "model" ARGS -m "${MODEL_DEST}")
+set_tests_properties(test-state-restore-fragmented PROPERTIES FIXTURES_REQUIRED test-download-model)
 
 if (NOT GGML_BACKEND_DL)
     # these tests use the backends directly and cannot be built with dynamic loading