[ggml] Add ggml linear q4_0

Summary:

Test Plan:

Reviewers:

Subscribers:

Tasks:

Tags:

ghstack-source-id: 641e19b
Pull Request resolved: #2385

Author: larryliu0820

.gitmodules

@@ -59,3 +59,6 @@
 	path = third-party/lm-evaluation-harness
 	url = https://github.com/EleutherAI/lm-evaluation-harness
 	branch = v0.4.1
+[submodule "examples/third-party/llama.cpp"]
+	path = examples/third-party/llama.cpp
+	url = https://github.com/ggerganov/llama.cpp.git

examples/models/llama2/playground/CMakeLists.txt

@@ -0,0 +1,68 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+#
+# Simple CMake build system for selective build demo.
+#
+# ### Editing this file ###
+#
+# This file should be formatted with
+# ~~~
+# cmake-format --first-comment-is-literal=True CMakeLists.txt
+# ~~~
+# It should also be cmake-lint clean.
+#
+
+cmake_minimum_required(VERSION 3.19)
+project(QuantizedLinearOp)
+if(NOT CMAKE_CXX_STANDARD)
+  set(CMAKE_CXX_STANDARD 17)
+endif()
+
+set(EXECUTORCH_ROOT ${CMAKE_CURRENT_SOURCE_DIR}/../../../..)
+set(TORCH_ROOT ${EXECUTORCH_ROOT}/third-party/pytorch)
+
+set(_common_compile_options -Wno-deprecated-declarations -fPIC)
+
+# Let files say "include <executorch/path/to/header.h>".
+set(_common_include_directories ${EXECUTORCH_ROOT}/..)
+
+find_package(Llama REQUIRED)
+find_package(ExecuTorch REQUIRED)
+find_package(Torch CONFIG REQUIRED)
+
+target_include_directories(executorch INTERFACE ${_common_include_directories})
+
+
+set(kernel_sources ${EXECUTORCH_ROOT}/examples/models/llama2/playground/op_linear.cpp)
+#
+# custom_kernels: C++ kernel implementations of custom ops
+#
+add_library(custom_kernels SHARED ${kernel_sources})
+target_link_libraries(custom_kernels PRIVATE executorch ${LLAMA_LIBRARY})
+target_compile_options(custom_kernels PUBLIC ${_common_compile_options})
+target_include_directories(custom_kernels PRIVATE ${EXECUTORCH_ROOT}/examples/third-party ${TORCH_INCLUDE_DIRS})
+
+
+if(EXECUTORCH_BUILD_GTEST)
+    find_package(
+        gflags REQUIRED PATHS ${CMAKE_CURRENT_BINARY_DIR}/../../../../third-party
+    )
+    #
+    # llama_cpp_test: test binary to run llama.cpp kernel ggml_mul_mat
+    #
+    add_executable(llama_cpp_test PRIVATE ${EXECUTORCH_ROOT}/examples/llama2/playground/test_op_linear.cpp)
+
+    target_link_libraries(llama_cpp_test executorch gflags custom_kernels)
+    target_compile_options(llama_cpp_test PUBLIC ${_common_compile_options})
+endif()
+
+# Install libraries
+install(
+  TARGETS custom_kernels
+  DESTINATION lib
+  INCLUDES
+  DESTINATION ${_common_include_directories})