Leveraging the power of C++ for efficient machine learning on embedded devices Adrian Stanciu adrian.stanciu.pub@gmail.com CppCon, 2023 1 / 50About me ◮ I am a software engineer from Romania ◮ I have data to make predictions 11 / 50Neural network (NN) 13 / 50Convolutional neural network (CNN) ◮ Efficient in image classification ◮ A convolutional layer can apply filters to detect: ◮ Edges ◮ Shapes ◮ 50Conclusions ◮ Code isn’t enough... data matters ◮ More diverse data leads to better models ◮ Building accurate models is an expert job ◮ Running on-device inference is straightforward ◮ Running on-device

library aims to provide a simple yet powerful framework for building and training machine learning models. By leveraging the performance efficiency of C++, micro- grad++ offers a robust solution for integrating for defining layers and neurons, enabling users to construct complex network architec- tures. • Backpropagation: The implementation of backpropa- gation in micrograd++ allows for efficient training of of models through gradient descent. • Gradient Clipping: To prevent the issue of exploding gradients, micrograd++ incorporates gradient clipping, ensuring stable training processes. C. Unique Aspects and

Windows (none) AI contributions Contributions must not include content generated by large language models or other probabilistic tools, including but not limited to Copilot or ChatGPT. This policy covers contributions, something that we cannot trust an AI tool to do. Copyright issues: Publicly available models are trained on copyrighted content, both accidentally and intentionally, and their output often includes Ethical issues: AI tools require an unreasonable amount of energy and water to build and operate, their models are built with heavily exploited workers in unacceptable working conditions, and they are being used

Pages Controller::redirect() Forwarding to an Action on the Same Controller Loading Additional Models Controller::fetchModel() Controller::fetchTable() Paginating a Model Controller::paginate() Configuring Own Plugins Creating a Plugin Using Bake Plugin Classes Plugin Routes Plugin Controllers Plugin Models Plugin Templates Overriding Plugin Templates from Inside Your Application Plugin Assets Linking from users. It is responsible for rendering a response with the aid of both the Model and the View layers. A controller can be seen as a manager that ensures that all resources needed for completing a task

Redirecting to Other Pages Redirecting to Another Action on the Same Controller Loading Additional Models Paginating a Model Configuring Components to Load Configuring Helpers to Load Request Life-cycle Own Plugins Creating a Plugin Using Bake Plugin Objects Plugin Routes Plugin Controllers Plugin Models Plugin Templates Overriding Plugin Templates from Inside Your Application Plugin Assets Linking from users. It is responsible for rendering a response with the aid of both the Model and the View layers. A controller can be seen as a manager that ensures that all resources needed for completing a task

development tools In this section, I'm going to show you a few IDEs that can help you become a more efficient programmer, with capabilities such as intelligent code completion and auto-formatting. They are "Salut" // now the value of m["hello"] is Salut make, new make does memory allocation for built-in models, such as map , slice , and channel , while new is for types' memory allocation. new(T) allocates functions to operate on one variable. Low cost by passing memory addresses (8 bytes), copy is not an efficient way, both in terms of time and space, to pass variables. channel , slice and map are reference

app Write your first view Writing your first Django app, part 2 Database setup Creating models Activating models Playing with the API Introducing the Django Admin Writing your first Django app, part 3 mod_wsgi Get your database running Remove any old versions of Django Install the Django code Models and databases Models Making queries Aggregation Search Managers Performing raw SQL queries Database transactions Historical models Considerations when removing model fields Data Migrations Squashing migrations Serializing values Supporting multiple Django versions Managing files Using files in models The File object

app Write your first view Writing your first Django app, part 2 Database setup Creating models Activating models Playing with the API Introducing the Django Admin Writing your first Django app, part 3 mod_wsgi Get your database running Remove any old versions of Django Install the Django code Models and databases Models Making queries Aggregation Search Managers Performing raw SQL queries Database transactions Commands Backend Support Workflow Dependencies Migration files Adding migrations to apps Historical models Considerations when removing model fields Data Migrations Squashing migrations Serializing values

app Write your first view Writing your first Django app, part 2 Database setup Creating models Activating models Playing with the API Introducing the Django Admin Writing your first Django app, part 3 mod_wsgi Get your database running Remove any old versions of Django Install the Django code Models and databases Models Making queries Aggregation Search Managers Performing raw SQL queries Database transactions Commands Backend Support Workflow Dependencies Migration files Adding migrations to apps Historical models Considerations when removing model fields Data Migrations Squashing migrations Serializing values