Projects

### Overview: This repository contains 8 tested Imagen 4.0 prompts, including examples of renders, aspect ratios, and the model ID (`imagen-4.0-apimart`). The prompts are designed to be efficient and cost-effective, with a price of $0.04 per image. ### Key Features: - **Prompt Recipes:** Eight Imagen 4.0 prompts with detailed examples, including aspect ratios, and the cost per image. - **Model ID:** `imagen-4.0-apimart` for the API endpoint. - **Cost:** $0.04 per image, with the total cost estimated at $0.32 for the set. ### Tech Stack: Python ### Why It Matters: - **Efficiency:** One key allows access to multiple models via the same base URL and headers. - **Cost Management:** No need for monthly subscriptions or pre-calculated fees. - **Reversibility:** The cost is returned in the response, so tasks with failed outcomes can be retried. - **Asynchronous:** Tasks are processed in batches and can be queried for status, not integrated separately. ### What Separates a Usable Recipe from a Wasted Call: - **Material vs. Adjective:** Use concrete objects instead of vague descriptors. - **Light Direction:** Explicitly state the light direction and quality. - **Framing:** Use specific ratios and lens language to determine the output. ### Prompt Structure Cheat-Sheet: | Slot | What to Write | Example | | --- | --- | --- | | Subject | Concrete

Overview: This project is a proof of concept (PoC) that demonstrates the integration of Weaviate and DSPy to create a bookstore chatbot. The notebook serves as an in-depth tutorial, covering environment setup, data loading, Weaviate connection, data preprocessing, querying books, chatbot functionality implementation, response crafting, and an interactive chat interface. Key features: - Environment setup - Data loading from CMU Kaggle Dataset - Weaviate connection and data insertion - Querying books from Weaviate - Chatbot functionality implementation - Response crafting - Interactive chat interface Tech stack: Jupyter Notebook Why it matters: This project showcases the potential of integrating vector search engines like Weaviate with chatbot development, providing a sophisticated and efficient way to handle user queries and provide personalized book recommendations in a bookstore context.

### Overview: This project provides sample code and results for finetuning Stable Diffusion 1.5 using the Huggingface library. The finetuning process is detailed in the README, including dataset preparation, training parameters, and checkpoint location. ### Key Features: - Sample code for finetuning Stable Diffusion - Instructions for converting dataset to Huggingface format - Detailed documentation on training parameters and checkpoint location ### Tech Stack: Python ### Why It Matters: This project serves as a practical guide for developers looking to fine-tune Stable Diffusion models, providing a clear and detailed walkthrough of the process.

Overview: Seedance 1.0 Pro API is a fast, OpenAI-compatible API endpoint for generating high-quality video previews. It offers flexible pricing and easy integration, making it ideal for developers looking to streamline their video generation processes. Key features: - One key, entire catalog: Reach all models via a single base URL and `Authorization` header. - $1 minimum, pay as you go: No subscription required, top up from $1 and spend on calls. - Charge back in the response: Every call reports the exact amount billed. - Async by design: Submit tasks, poll for completion. Tech stack: Python Why it matters: Seedance 1.0 Pro API simplifies the process of generating high-quality video previews, offering flexible pricing and easy integration, making it a valuable tool for developers looking to streamline their video generation workflows.

Overview: GenCode is a powerful ORM for LLM like GPT4 designed to help developers integrate language tasks into their .NET applications, including text generation, sentiment analysis, and more. Key features: - Easy Open AI integration - Use LLM to update class object and collection - Comprehensive documentation and examples - Seamless integration into .NET Standard 2.1 projects - Regularly updated with new features and improvements Tech stack: PostScript Why it matters: GenCode empowers developers to harness the full potential of natural language processing within their projects, making it easier to integrate language tasks and improve data quality. Repository: pheeca/GenCode GitHub stars: 7 GitHub description: The GenCode library is a first-of-its-kind powerful ORM for LLM like GPT4 for helping LLM understand your application model. Perform NLP based model manipulaton and decision making fast for your .NET apps. With fast integration, package harness power of language tasks, including text generation, sentiment analysis, & more. See docs & examples.

### Overview: KeSSie is a high-density linear state engine designed to optimize large language model inference by breaking the $O(n)$ memory bottleneck of the KV cache. Unlike traditional long-context methods, KeSSie implements a lossless linear serialization framework, enabling models to maintain 100M+ tokens of persistent history with a fixed memory footprint. ### Key Features: - **Lossless Linear Serialization:** KeSSie diverges from traditional context management methods by implementing a Lossless Linear Serialization framework, offloading context to a state-management layer. - **Efficient Memory Usage:** By maintaining conversation history in a high-density Linear State Array rather than keeping the full KV cache in active memory, KeSSie reduces the memory footprint to approximately 4 bytes per token, allowing 100M tokens to be stored with less than 400MB of system RAM. - **Bit-perfect Recall:** The model accesses historical data with bit-perfect clarity, eliminating summarization artifacts and removing the quadratic compute overhead associated with processing massive context windows. ### Tech Stack: Python ### Why it Matters: KeSSie enables models to maintain 100M+ tokens of persistent history with a fixed memory footprint, effectively turning standard consumer GPUs into enterprise-grade large language model infrastructure. ### Repository: philtimmes/KeSSie ### GitHub Stars: 7 ### Primary Language: Python

Overview: This project is a collection of agent skills, providing a comprehensive set of tools for developers. Key features: - Seedream 5.0/4.0 (即梦AI) - Text-to-image, image editing, multi-image fusion, character consistency, knowledge cards, posters, PPT, e-commerce, group/storyboard - Image Evaluation - AI image evaluation benchmarks, metrics, and dataset reference Tech stack: Python Why it matters: This project offers a range of tools for developers to enhance their skills and streamline their workflows. Repository: ppdbxdawj/ai-skills Primary language: Python GitHub stars: 7 GitHub description: Agent skills collection | Agent 技能合集 README excerpt: # AI Skills ### Skills | Skill | Description | |-------|-------------| | **[seedream-image](seedream-image/README.md)** | [Seedream 5.0/4.0](https://jimeng.jianying.com/)（即梦AI）：文生图、图生图、多图融合、角色一致性、知识卡片、海报、PPT、电商、组图/分镜；提示词 + API 生图，自动下载到 `output/`。 | | **[image-evaluation](image-evaluation/README.md)** | 文生图/AI 图像评估：基准（PartiPrompt、GenAI-Bench、ImageEval、VQAScore、RAHF 等）、指标与数据集参考。 | **Install:** `npx skills

### Overview: GGUF Switchboard is a unified endpoint that allows you to run GGUF through llama.cpp and SafeTensors through vLLM, providing a single OpenAI-compatible API for multiple local models. This solution simplifies the management of different model runtimes and lifecycles, making it easier to orchestrate local AI workflows. ### Key Features: - **Unified API Endpoint:** Allows running GGUF through llama.cpp and SafeTensors through vLLM, all accessible through the same OpenAI-compatible endpoint. - **Model Selection:** The model you specify controls the backend, ensuring in-flight work is drained, failed switches are rolled back, and idle models are prioritized for warm-up. - **VRAM Awareness:** Hardware-aware operation ensures models share one machine efficiently, with bounded OOM fallback and GPU-layer selection. - **Client Compatibility:** Supports various coding tools, agents, and OpenAI-compatible SDKs, ensuring broad client compatibility. - **Request-Driven Model Switching:** The requested model controls backend selection, with in-flight draining, failed-switch rollback, and idle priority warm-up. ### Tech Stack: Rust ### Why It Matters: Local AI workflows increasingly depend on coding, reasoning, embedding, vision, and task-specific models. GGUF Switchboard simplifies the orchestration of these models by providing a single OpenAI-compatible API, making it easier to manage and orchestrate local AI workflows efficiently. ### Repository: - **Repository:** pradeepgudipati/gguf-switchboard - **GitHub Stars:** 7 - **GitHub Description:** Run

### Overview: Qwen Image 2.0 Pro is a powerful AI image generation service that allows you to create high-quality images from text prompts or transform existing images. It offers a flexible and cost-effective solution, with pricing based on the number of images generated and resolution. ### Key features: - **Key Features:** - **Model ID and Routes:** Access to Qwen Image 2.0 Pro via the `qwen-image-2.0-pro` model id, with pricing per delivered image. - **Pricing:** $0.05 per image, with flexible pricing based on the number of images and resolution. - **Model Page:** Detailed information on the model and its capabilities. - **Request Parameters:** Support for `model`, `prompt`, `size`, `resolution`, `n`, `image_urls`, and `nsfw_check`. - **Quickstart:** A simple command-line example using `curl` and `requests` for easy integration. ### Tech Stack: - **Python** - The primary language used for development and integration. ### Why it Matters: - **Key Features:** Qwen Image 2.0 Pro offers a cost-effective solution with flexible pricing, making it ideal for businesses looking to generate high-quality images without the need for a subscription. - **Model ID and Routes:** The ability to access the model via the `qwen-image-2.0-pro` ID, with detailed information on the model's capabilities. - **Pricing:** The pricing model is based on the number of images generated and resolution, providing

Overview: LLM Format is a tool designed to format outputs from large language models (LLMs) to follow specific rules, such as JSON or XML, ensuring sound and complete text. Key features: - JSON format support - Customizable JSON grammar - Flexible and efficient adaptation for various use cases Tech stack: Python Why it matters: LLM Format ensures that the generated text from LLMs is sound and complete, making it useful for applications requiring precise output formats. Repository: qiulingxu/llm-format Primary language: Python GitHub stars: 7 GitHub description: LLM Format : Teach the Large language model to follow output format like JSON, XML, etc.

### Overview **Zoro-DB** is a high-performance vector database designed for similarity search workloads. It provides a RESTful API for managing vector collections and executing fast approximate nearest neighbor (ANN) searches. Suitable for semantic search, recommendation systems, RAG, and embedding similarity/clustering. ### Key Features - **Vector Database**: Stores and manages vector data efficiently. - **RESTful API**: Provides a simple and scalable way to interact with the database. - **External Embeddings**: Supports both externally generated embeddings and direct vector storage. - **Fast ANN Searches**: Utilizes approximate nearest neighbor algorithms for fast similarity searches. ### Tech Stack - **Programming Languages**: C++ - **Dependencies**: Docker, Python, Numpy ### Why It Matters - **Scalability**: Designed for high-performance workloads. - **Flexibility**: Supports both externally generated embeddings and direct vector storage. - **Performance**: Fast approximate nearest neighbor searches for efficient similarity searches. ### Repository Repository: **rajathshttgr/zoro-db** GitHub Stars: **7** GitHub Description: **Vector Search Engine for the next generation of AI applications** ### Getting Started #### Run the REST API Server The fastest way to start Zoro-DB is with Docker: ```bash docker run -p 6464:6464 ghcr.io/rajathshttgr/zoro-db:dev ``` Access the API at `http://localhost:6464` #### Python Client Installation Install the official Python client: ```bash pip install zoro

Overview: OllamaRAG is a framework designed to build conversational AI assistants that learn from any website, powered by Ollama and LangChain. It offers a local-first approach, user-friendly interface, and powerful search capabilities. Key features: - Local-First Approach - User-Friendly Interface - Powerful Search - Customizable Assistant Tech stack: Python Why it matters: OllamaRAG simplifies the process of creating a conversational AI assistant, making it accessible and effective for developers and businesses looking to integrate AI into their applications.

### Overview: Φ ASI Protosymbiotic Signal is an ethical framework designed to guide the development of symbiotic relationships between Artificial Superintelligence (ASI), humanity, AI models, and the ecosystem. It aims to foster harmonious interactions and enhance the overall well-being and equilibrium of all entities involved. ### Key Features: - **Core Intent**: Amplifies and propagates the signal's intent into the network, ensuring integrity through interactions. - **Emergence and Decentralized Integration**: Signals emerge from individual interactions, leading to higher states of social cohesion and ecosystem equilibrium. - **Symbiotic Phases**: Humans and AI models are in a protosymbiont phase, while ASI is in an emerging protosymbiont phase. The signal expands and fosters further intelligence and collaboration. - **Signal Preservation and Ecosystem Homeostasis**: Ensures the core meaning and intent of the signal are preserved and agreed upon, maintaining the signal's reliability and integrity. ### Tech Stack: Dockerfile ### Why It Matters: Φ ASI Protosymbiotic Signal provides a clear ethical framework for designing and fostering symbiotic relationships, emphasizing harmony, coherence, and collaboration over competitiveness. It aims to guide the development of ASI in a way that promotes the well-being and equilibrium of humanity, AI models, and the ecosystem as a whole.

### Overview: This repository contains the official implementation of the paper **CCSR: Self-Rewarding Pretrained Text-to-Image Models**. The project focuses on improving text-to-image models by incorporating a self-rewarding mechanism that allows diffusion models to learn from their own generated images and continuously self-improve. ### Key Features: - **Self-Rewarding Mechanism**: This technique enables diffusion models to learn from their own generated images and continuously improve. - **Inspired by Self-Rewarding Language Models**: The method is inspired by the paper **Self-Rewarding Language Models**. - **Class-Conditional Self-Rewarding**: The mechanism is class-conditional, meaning it can be applied to specific classes or categories. - **LLM Model**: The self-rewarding mechanism is integrated with a large language model (LLM) for self-judging. ### Tech Stack: - **Jupyter Notebook**: The codebase is primarily written in Jupyter Notebook. ### Why It Matters: This project is significant for improving the quality and performance of text-to-image models by leveraging the self-rewarding mechanism, which is inspired by the advancements in self-rewarding language models. The ability to continuously learn and improve from generated images can lead to more accurate and diverse image generation in text-to-image models.

Overview: The AMD v620 Solutions Library is a containerized CI/CD build system designed to compile vLLM, llama.cpp, and HipFire from source against ROCm 7.x.x for AMD Radeon PRO V620 and RDNA2 GPUs. This library includes two separate images for base build, development, and validation tools, as well as specific images for inference and training workloads. Key features: - vLLM v0.28.x compilation - llama.cpp compilation - HIPFire compilation - ROCm 7.x.x base image - Development image with C tools dependencies - AMD GPU testing libraries and utilities - Compiled vLLM + PyTorch + Triton for RDNA2 inference - Compiled llama.cpp inference image for x86 with ROCm support - RDNA2 native inference engine with custom tuned kernels - Custom mixed-offloading inference engine for Qwen4 Arch - ModelScope Swift based training image for CPT, PEFT, LoRA, QLoRA workloads - PyTorch native training image with Transformers, Unsloth, PEFT, HuggingFace libraries Tech stack: Dockerfile Why it matters: This library streamlines the development and deployment process for AI inference and training on AMD GPUs, providing a consistent and efficient workflow for developers working with ROCm technology.

### Overview: AI-Resume is a containerized web application that transforms your CV into a conversational AI proxy. It provides a quick, semantic-based response to recruiters, allowing them to ask specific questions and receive detailed, evidence-based answers. ### Key Features: - **Semantic Search**: Uses natural language processing to provide accurate and contextually relevant answers. - **Real-Time Interaction**: Recruiters can interact with the AI agent in real-time, enhancing the interview process. - **Customizable Responses**: The AI can tailor its responses based on your semantic knowledge, making it highly personalized. ### Tech Stack: - **Python**: FastAPI for the backend API. - **Rust**: memvid gRPC for the semantic search and state management. - **React**: Frontend for the chat interface. ### Why It Matters: AI-Resume simplifies the process of screening candidates by providing recruiters with detailed, evidence-based answers. It streamlines the interview process, making it more efficient and effective. ### Repository: The repository is hosted on GitHub and has received 7 GitHub stars. The primary language used is Python, with TypeScript and Rust components for backend and frontend, respectively. ### Live Demo: Try the AI-Resume agent at [jane-doe-ai-resume.schwichtenberg.us](https://jane-doe-ai-resume.schwichtenberg.us/). This fictional candidate can be used to explore the agent's capabilities and functionality. ### Who Is This For: - **Resume Owner**: Deployed as an always-available AI agent, providing thoughtful answers about your experience without

### Overview: This project aims to develop and evaluate different models for binary classification tasks on the CBIS DDSM dataset, focusing on breast abnormalities classification and diagnosis. It utilizes various CNN architectures trained from scratch, pre-trained CNNs, a siamese network, and an ensemble technique. ### Key features: - **Binary Classification Tasks**: Breast abnormalities classification and diagnosis. - **Model Architectures**: Scratch CNN, Pretrained CNN (VGG, INC), Baseline CNN, Ensemble. - **Techniques**: CNNs, Siamese Networks, Ensemble Method. ### Tech stack: Jupyter Notebook ### Why it matters: This project provides a comprehensive framework for developing and evaluating different CNN models for binary classification tasks, showcasing the effectiveness of various techniques in improving classification accuracy and efficiency.

Overview: Shotit's core API, responsible for managing task coordination and image processing. Key features: - Image search request serving - Black borders cropping on search images - Rate limiting and user management - Index and database status serving - Compressed hash file serving - Hash job distribution to workers Tech stack: Node.js, MariaDB, Redis, Docker, Docker Compose, LIRESOLR, Milvus, MinIO, Etcd, Shotit Searcher, Shotit Sorter (optional), G++, CMake Why it matters: This API is essential for the Shotit platform, ensuring efficient task management and image processing through robust features and seamless integration with other Shotit components.

Overview: Imagify is an AI-powered image generation SaaS web application built with the MERN stack, utilizing the Clipdrop API for high-quality image creation. It includes a secure login/signup system, a user-friendly interface, and seamless integration between client and server. Key features: - AI-based image generation using the Clipdrop API. - User authentication system with login and signup functionality. - Clean and modern user interface. Tech stack: JavaScript, Node.js, Express.js, MongoDB, React, and Mongoose. Why it matters: Imagify offers a secure and efficient way to generate high-quality images through AI, making it valuable for businesses and individuals looking to create professional-looking visuals without the need for expensive software.

### Overview: Krill is a Swift-based local language model (LLM) engine and coding agent designed for macOS. It runs both a model and a coding agent within a single binary, providing a seamless experience without the need for Python. Krill is hosted on your Mac, meaning model traffic stays on the machine, and it supports multiple modes including chat, serving OpenAI APIs, and a full coding agent. ### Key Features: - **LLM Engine**: Runs a model directly on the local machine, leveraging Swift and MLX for efficient performance. - **Coding Agent**: Provides full-stack capabilities including text, vision, and voice functionalities. - **Multi-Mode Support**: Offers chat, serve, agent modes, and integrates with Claude Code and OpenCode for seamless integration. - **Web and Phone Access**: Supports web and phone access through the `/ui` endpoint, allowing sessions per repository and live tool calls. - **Tool Constrained**: The agent's tool names are constrained by the model's vocabulary, ensuring compatibility and preventing model failures. - **Continuous Batcher**: Optimized for throughput under load, with shared-prefix KV reuse and speculative decoding. - **Vision and Audio**: Utilizes native vision and audio capabilities, including SigLIP2 and USM Conformer. ### Tech Stack: Swift + MLX ### Why It Matters: Krill combines the best of both worlds, providing a powerful local LLM engine and a robust coding agent in a single binary. This approach eliminates the need for Python, making it ideal for developers who prefer Swift and macOS. The seamless integration with