Search Results for "foo_input_dvda.fb2k-component"

...The project emphasizes portability and performance, enabling machine learning inference across a wide range of hardware environments including CPUs and specialized accelerators. It is widely used as a foundational component in projects that run large language models locally, including tools that perform inference for transformer-based models. The library also implements optimization algorithms and computation graph functionality so developers can build training and inference workflows directly on top of its tensor operations.

COCOON is a privacy-aware desktop client framework designed by the developers of Telegram to provide a modern, secure, and extensible environment for building messaging and communication applications. At its core, it combines native desktop performance with web-like flexibility, packing a renderer, UI components, and plugin architecture that allows developers to craft rich experiences similar to those found in native apps. Cocoon’s architecture prioritizes privacy and security, making it...

...Most notably, it contains a lightning-fast "fully fused" multi-layer perceptron (technical paper), a versatile multiresolution hash encoding (technical paper), as well as support for various other input encodings, losses, and optimizers. We provide a sample application where an image function (x,y) -> (R,G,B) is learned. The fully fused MLP component of this framework requires a very large amount of shared memory in its default configuration. It will likely only work on an RTX 3090, an RTX 2080 Ti, or high-end enterprise GPUs. Lower-end cards must reduce the n_neurons parameter or use the CutlassMLP (better compatibility but slower) instead. tiny-cuda-nn comes with a PyTorch extension that allows using the fast MLPs and input encodings from within a Python context. ...

...So I decided to write a super small and hackable inference library specifically focused on minimizing memory consumption: OnnxStream. OnnxStream is based on the idea of decoupling the inference engine from the component responsible for providing the model weights, which is a class derived from WeightsProvider. A WeightsProvider specialization can implement any type of loading, caching, and prefetching of the model parameters.

Recast Navigation is an open source toolkit designed for building navigation meshes and performing pathfinding in 3D environments, particularly for video games and simulations. It provides a set of libraries that transform input geometry into navigation meshes that AI agents can use to move through virtual worlds. Its core component, Recast, constructs navigation meshes by rasterizing triangle meshes into voxels, filtering out areas that are not walkable, and converting the remaining regions into polygon meshes suitable for navigation. Detour complements this system by providing runtime pathfinding, spatial queries, and navigation utilities that allow agents to move efficiently across the generated mesh. ...

MLPACK is a C++ machine learning library with emphasis on scalability, speed, and ease-of-use. Its aim is to make machine learning possible for novice users by means of a simple, consistent API, while simultaneously exploiting C++ language features to provide maximum performance and flexibility for expert users. * More info + downloads: https://mlpack.org * Git repo: https://github.com/mlpack/mlpack

...In a single repository, Flashlight provides apps for research across multiple domains. Flashlight can be broken down into several components as described above. Each component can be incrementally built by specifying the correct build options. Flashlight is most-easily built and installed with vcpkg. Both the CUDA and CPU backends are supported with vcpkg. For either backend, first, install Intel MKL. Flashlight app binaries are also built for the selected features and are installed into the vcpkg install tree's tools directory.

OSTIS - Open Semantic Technology for Intelligent Systems. This is an open project focused on mass semantic technology creation used in component-based design of various intelligent systems.

RoboComp is a robotics framework providing a set of open-source, distributed, real-time robotic and artificial vision software components and the necessary tools to create and manage them.

CRF decoder is the simplified version of CRF++, only for decoding the sequential data. It removes the training component and its correspondent codes from CRF++, which makes CRF decoder more reabable and understandable for freshman.

Orca is a component-based robotic framework. Its aim is to allow creation of complex robotic systems, from single vehicle architectures to distributed sensor networks, by easily connecting new components to reusable components from the Orca repository.

PlexBench is a cross-platform, web-enabled, analysis tool that is driven by a scalable backpropagation feed-forward neural network. It uses embedded Perl for scripting and is written in the style of an in-process Component Object Model (COM) C++ program.

Blackboard implements a lightweight, portable tuple space suitable for multi-agent system and distributed component design. Supports implicit invocation via content-filtered asynchronous events, blocking call semantics, and shared memory messaging.

This project aims at developing a face authentication system, using the Eigenfaces, and Eigenfeatures. The Approach is a principal component analysis method, in which a set of characteristic pictures are used to describe the variation between face images.

3D mobile robot simulation environment. Due to the component-based design very flexible and also suitable for several other mechatronical systems. Planned to be used for a real distributed robot system currently being built up.