script: refine script according to https://github.com/quic/ai-hub-apps/tree/main/tutorials/llm_on_genie

Author: jeffzhou2000

CMakeLists.txt

@@ -8,15 +8,16 @@ set(CMAKE_WARN_UNUSED_CLI YES)
 set(CMAKE_EXPORT_COMPILE_COMMANDS ON)
 
 if(CMAKE_SYSTEM_NAME STREQUAL "Android")
+    set(CMAKE_VERBOSE_MAKEFILE      ON)
     if(DEFINED HTP_ARCH_VERSION)
         if (${HTP_ARCH_VERSION} STREQUAL "v75" OR ${HTP_ARCH_VERSION} STREQUAL "v79")
             #works fine on Snapdragon 8Gen3&8Elite with 1.5x - 3x performance gains with the default ggml backend
             set(OPT_FLAG " -O3 -march=armv8.7-a -mcpu=cortex-x1 -mtune=cortex-x1 -ffp-model=fast -fno-finite-math-only")
             message("OPT_FLAG:${OPT_FLAG}")
-            set(CMAKE_C_FLAGS   "${CMAKE_C_FLAGS}   -DGGML_USE_HEXAGON ${DEBUG_FLAG} ${OPT_FLAG}")
-            set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DGGML_USE_HEXAGON ${DEBUG_FLAG} ${OPT_FLAG}")
-            set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -DGGML_USE_HEXAGON ${DEBUG_FLAG} ${OPT_FLAG}")
-            set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -DGGML_USE_HEXAGON ${DEBUG_FLAG} ${OPT_FLAG}")
+            set(CMAKE_C_FLAGS   "${CMAKE_C_FLAGS}   -DGGML_USE_HEXAGON -DGGML_USE_LLAMAFILE ${DEBUG_FLAG} ${OPT_FLAG}")
+            set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DGGML_USE_HEXAGON -DGGML_USE_LLAMAFILE ${DEBUG_FLAG} ${OPT_FLAG}")
+            set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -DGGML_USE_HEXAGON -DGGML_USE_LLAMAFILE ${DEBUG_FLAG} ${OPT_FLAG}")
+            set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -DGGML_USE_HEXAGON -DGGML_USE_LLAMAFILE ${DEBUG_FLAG} ${OPT_FLAG}")
         endif()
     endif()
 endif()

ggml/src/ggml-hexagon/CMakeLists.txt

@@ -39,7 +39,9 @@ endif()
 
 #check whether user's specified htp arch is valid
 set(CHECK_HTP_ARCH "WRONG")
-foreach (feat v68 v69 v73 v75 v79)
+#ref: https://github.com/quic/ai-hub-apps/tree/main/tutorials/llm_on_genie
+#foreach (feat v68 v69 v73 v75 v79)
+foreach (feat v73 v75 v79)
     if (${feat} STREQUAL ${HTP_ARCH_VERSION})
         set(CHECK_HTP_ARCH "GOOD")
     endif()

scripts/build-run-android.sh

@@ -61,23 +61,24 @@ PROMPT_STRING="every day of your life, it is important to take the time to smell
 PROMPT_STRING="introduce the movie Once Upon a Time in America briefly.\n"
 
 #for llama-cli, 20.4 MiB in models/t5-very-small-random-F32.gguf
-TEST_MODEL_NAME=/sdcard/t5-very-small-random-F32.gguf
+#TEST_MODEL_NAME=/sdcard/t5-very-small-random-F32.gguf
 #for llama-cli, 1.1 GiB, will be downloaded automatically via this script
 #TEST_MODEL_NAME=/sdcard/t5-277M-F32.gguf
+TEST_MODEL_NAME=/sdcard/qwen1_5-1_8b-chat-q4_0.gguf
+#self-defined LLM models
+#TEST_MODEL_NAME=/sdcard/Qwen3-8B-Q8_0.gguf
+#TEST_MODEL_NAME=/sdcard/Qwen3-4B-Q8_0.gguf
+#TEST_MODEL_NAME=/sdcard/gemma-3-4b-it-Q8_0.gguf
+
 
 #for llama-bench, 1.12 GiB, will be downloadded automatically via this script
 GGUF_MODEL_NAME=/sdcard/qwen1_5-1_8b-chat-q4_0.gguf
 
-#available htp arch version:
-#v68 --- Snapdragon 888
-#v69 --- Snapdragon 8 Gen1
+#ref: https://github.com/quic/ai-hub-apps/tree/main/tutorials/llm_on_genie
+#supported htp arch version:
 #v73 --- Snapdragon 8 Gen2
 #v75 --- Snapdragon 8 Gen3
-#v79 --- Snapdragon 8 Elite(aka Gen4)
-
-#8Gen1
-#HTP_ARCH_VERSION=v69
-#HTP_ARCH_VERSION_a=V69
+#v79 --- Snapdragon 8 Elite
 
 #8Gen2
 #HTP_ARCH_VERSION=v73