Quantcast
Channel: Raspberry Pi Forums
Viewing all articles
Browse latest Browse all 4919

SDK • Re: Best Practices for Using TensorFlow Lite for Microcontrollers as a Library

$
0
0
I guess I was overcomplicating things.

I did not need to prepare any intermediate CMakeLists.txt file, and I was able to build only the code I needed without any problems by simply writing the following in my project's CMakeLists.txt file. All I needed was this.
add_subdirectory(vendor/pico-tflmicro EXCLUDE_FROM_ALL)

Code:

cmake_minimum_required(VERSION 3.13...3.27)include(vendor/pico_sdk_import.cmake)project(pico-neuralclick C CXX ASM)set(CMAKE_C_STANDARD 11)set(CMAKE_CXX_STANDARD 11)pico_sdk_init()# include target library pico-tflmicroadd_subdirectory(vendor/pico-tflmicro EXCLUDE_FROM_ALL)add_executable(neuralclick main.cpp)target_link_libraries(neuralclick PRIVATE  hardware_sync  pico_stdlib  pico-tflmicro)pico_enable_stdio_usb(neuralclick 1)pico_add_extra_outputs(neuralclick)
Now I can safely use pico-tflmicro for projects involving machine learning.

Thank you.

Statistics: Posted by 0yama — Fri Jun 28, 2024 3:14 am



Viewing all articles
Browse latest Browse all 4919

Trending Articles