Using this library the developer can acceralarte the task of inference by liveraging the Nvidia GPU, without worrying about writing the GPU specific code.
- Nvidia GPU
- CUDA ( preferabbly latest version )
The repo has the following projects
| Project name | Description |
|---|---|
| DeepLearningInference | The source-code for the Deep learning Inference library |
| Sample Dense network | Sample project to demo Vanilla Dense NN |
| Sample ConvNet | Sample project to demo Vanilla CNN |
- Add support for Conv and Pool layers.
- Add Logger
- Document How to load weights ?
- Before that, provide a clean, consistent and scalable way to load the weights without any dependency on NN architecture.
- Refactor the code for better readility and maintainability by following the C++ best-practices and STLs.
- Optimize the code for efficient use of Device and Host Memory
- Check if we can use Thrust, the CUDA C++ template library
- Add Unit-tests and Assert statements whereever necessary.
- Brainstorm on the pros and cons(in terms of sacalability, maintainability and ease of use) of having a ContextFactory. If possible, get ride of the ContextFactory.