Search Results for "network fast"

Needle is an experimental 26-million-parameter function-calling model designed to run on extremely small devices such as phones, watches, glasses, and low-power personal AI hardware. It is based on a Simple Attention Network architecture and was distilled from a much larger model to focus on fast, compact tool-use behavior. The project provides open weights, training details, dataset generation resources, and a playground for testing the model with custom tools. Needle is optimized for single-shot function calling rather than broad conversational ability, so its core use case is selecting the right tool and producing structured arguments. ...

SlowFast is a video understanding framework that captures both spatial semantics and temporal dynamics efficiently by processing video frames at two different temporal resolutions. The slow pathway encodes semantic context by sampling frames sparsely, while the fast pathway captures motion and fine temporal cues by operating on densely sampled frames with fewer channels. Together, these two pathways complement each other, allowing the network to model both appearance and motion without excessive computational cost. The architecture is modular and supports tasks like action recognition, temporal localization, and video segmentation, performing strongly on benchmarks like Kinetics and AVA. ...