PGTFormer is an advanced video face restoration framework that restores high-fidelity details in videos while enhancing temporal coherence through a parsed guided temporal consistency transform. This method requires no pre-alignment and selects the most...
HivisionIDPhoto is a lightweight AI-powered ID photo maker that intelligently recognizes and cuts out backgrounds, quickly generating ID photos that meet various specifications. It supports custom background colors and sizes, and will introduce beauty filters and more in the future...
Step-1X is a large-scale AI image generation model launched by Step-1X Technology. It employs a self-developed DiT architecture and excels in deep semantic understanding and detail generation. Step-1X supports complex commands of up to 2000 characters, accurately matching text and images, and is suitable for...
Melty is an open-source AI programming assistant designed to improve developers' coding efficiency and code quality. Melty understands developers' programming activities in real time, from terminal operations to GitHub interactions, providing intelligent collaboration and...
Yi-Coder is an open-source programming assistant model series launched by Zero-One-Way, designed to improve the efficiency of tasks such as code generation, understanding, debugging, and completion. The Yi-Coder series includes versions with two parameter scales: 1.5B and 9B...
VideoGameBunny (VGB) is an open-source, large-scale multimodal model designed specifically for video games, developed by a research team at the University of Alberta, Canada. It can understand and generate game-related content in multiple languages and supports highly customizable...
LanceDB is a serverless vector database designed specifically for artificial intelligence applications. It supports vector search, full-text search, and SQL queries, and is optimized for multimodal data processing. It employs vector indexing technology to achieve efficient similarity search...
NNanoFlow is a high-performance service framework designed for large language models (LLMs) to improve inference throughput. It optimizes model performance by using computation, memory, and network resources in parallel within a single device...
ChatLearn is a flexible, easy-to-use, and efficient large-scale alignment training framework launched by Alibaba Cloud. ChatLearn is designed to support alignment training of large language models (LLMs). ChatLearn...
Docmatix is a large-scale dataset designed for Document Visual Question Answering (DocVQA) tasks. It contains 2.4 million images and 9.5 million question-answer pairs...
Kotaemon is an open-source tool based on RAG technology that allows users to interact with documents using natural language, quickly retrieving and understanding information. It is suitable for situations requiring the processing of large volumes of documents, such as academic research and enterprise document management...
The Stable Diffusion integration package is a local deployment solution for SD painting released by Bilibili UP master Qiuye aaaki. It includes the Stable Diffusion WebUI, the necessary runtime environment, pre-installed models, and a collection of commonly used plugins...
MiniCPM 3.0 is a high-performance edge AI model launched by Wallfacer AI. It features 4B parameters and achieves performance exceeding GPT-3.5 with a relatively small model size. MiniCPM 3.0 utilizes LLMxMapReduce technology and supports...
Reflection 70B is an open-source AI model from HyperWrite that outperforms GPT-4o and Llama 3.1 in multiple benchmark tests with its superior performance. Reflection 70B is based on Meta's Llama 3.1 70B Instruments...
Loopy is an audio-driven AI video generation model launched by ByteDance. Users can animate a still photo, with the facial expressions and head movements of the people in the photo synchronized with a given audio file to generate realistic videos...
xLAM is a large, open-source language model from Salesforce, designed specifically for function call tasks. The model can understand and execute API calls based on natural language instructions, and it excels in automating tasks and interacting with various digital services...
CodeFormer is an AI-powered photo and video restoration tool jointly developed by Nanyang Technological University and SenseTime. It integrates Variational Autoencoder (VQGAN) and Transformer technology to restore blurred and pixelated photos or videos...
Mini-Omni is an open-source, end-to-end voice dialogue model with real-time voice input and output capabilities, enabling "thinking while speaking" in conversations. The model is designed to support dialogue without requiring additional automatic speech recognition...
Composio is an AI agent development tool that provides over 100 integrated tools, simplifying the development and deployment of AI agents. It allows developers to call multiple tools and frameworks, such as OpenAI and Claude, with a single line of code...
DeepSeek-V2.5 is a new open-source model from DeepSeek that integrates general-purpose and code-based capabilities. It retains the general dialogue capabilities of the previous Chat model and the code processing capabilities of the Coder model, better aligning with human...
MLE-Agent is an intelligent assistant designed for machine learning engineers and researchers. It automates baseline creation, integrates the latest research resources, provides intelligent debugging, integrates file systems and tools, and offers interactive command-line chat...
ViewCrafter is an advanced video diffusion model jointly proposed by Peking University, the Chinese University of Hong Kong, and Tencent. It can synthesize high-fidelity new views from single or small numbers of images. It combines the generative capabilities of video diffusion models with point-based 3D representations...
FluxMusic is an open-source music generation model that uses a diffusion model and Transformer architecture to convert text descriptions into music. The model can handle complex text instructions and generate music with specific emotions, styles, and instrumentation...
LightEval, developed by Hugging Face, is a lightweight AI evaluation tool specifically designed for evaluating large language models (LLMs). LightEval supports multitasking and complex model configurations, runs on various hardware, and includes...