VideoMind is a Chain-of-LoRA Agent designed for long video reasoning using human-like processes.
AI模型接口管理与分发系统，支持多种大模型统一调用，并提供企业和个人使用的分发管理服务。
Code for Segment Any Motion in Videos, enabling motion segmentation in video sequences.
A Python-based automated testing framework for evaluating the performance and inference capabilities of large language models.
KIMI AI is a long-text model reverse API supporting high-speed streaming output, intelligent dialogue, and document interpretation.
Run LLMs with MLX, a Python package for generating text and fine-tuning large language models on Apple silicon.
Official code for the CVPR25 oral paper on biomechanically accurate human reconstruction.
Implementation of all RL algorithms in a simpler way - FareedKhan-dev/all-rl-algorithms
Train your AI self, amplify you, bridge the world with the Second Me open-source prototype.
A fork of llama.cpp with enhancements for performance and state-of-the-art quantization methods.
OpenDeepSearch is a lightweight, open-source search tool enabling deep web search with AI agent integration.
Dolphin is a multilingual, multitask ASR model jointly trained by DataoceanAI and Tsinghua University.
