Ian Villanueva@sprites20847

The system generates emotions from images, primarily intended for game NPCs so that they would react emotionally to the game environment. The sentiment analysis model is an LSTM-based Dense Neural Network that are fed Word2Vec embeddings. The model was trained using generated data from cohere.ai using the prompt: "I felt <emotion> when I saw <img2text>" System Flow: img2txt -> cohere.ai generator -> text2emote -> LSTM+DenseNN The images are passed to a CLIP Interrogator (BLIP + CLIP (ViT-32-B)) to generate text descriptions. Such text descriptions are elaborated by cohere.ai generator to generate emotional responses using the prompts: "When I saw <img2txt output>, I felt emotions such as"

(Work in progress) Please go to the GitHub page. The presentation link is troll since I can't make a for such a web app for such a short time. Soon I would implement agent creation and the actual Yi model. TY! In the realm of artificial intelligence, our project stands as a beacon of innovation, set to redefine machine capabilities. We aspire to create an AI system that not only automates tasks but comprehends the nuances of human thought. By integrating cloud computing and edge devices, unifying language models, and incorporating autonomous learning, we envision a future where AI is an intuitive companion. Cloud computing offers computational prowess, while edge devices provide local agility. Unifying language models with macros, interpreters, and the Retrieval-Augmented Generation (RAG) approach enhances linguistic capabilities. Our AI evolves autonomously, generating programs and refining itself. The YI model, with a colossal context size, enables nuanced responses. Join us on GitHub to witness AI's future.

Code: https://github.com/sprites20/Anthroid-AI Using together.ai to host the LLM serverless, Vectara for querying the documents, LLamaIndex for text embeddings (or LLM), and unstructured.io cleaning and translating the HTML or maybe even formatting the prompts. Uses Google Search API or Azure Bing Search service API to query the search engine, returns links, and sends to Vectara for indexing and querying, (perhaps more options in the future). For now, it uses only Vectara but will implement the LLamaindex soon. Shall also use the Meta's Graph API to gather posts, not only from search engines but social media apps like Facebook for the latest news about a topic using its query engine and more content not only from websites but from actual people in real-time. (WIP) It is also capable of choosing, retrieving code, and running code. With a built-in Python interpreter; the exec() function, that can also run other languages via bindings like jnius (Java), Cython/CPython (C/C++), C# DLLs, and whatever binding in the Python library there is. Even OpenGL for 3D rendering for true multimodality, OpenCV for RTSP streaming, image processing, and computer vision, matplotlib for mathematical visualizations, or even a custom web browser like Chrome and Edge that can also run JS evaluate to execute JS code for websites. A cross-platform native app, that in the future should be able to run on most operating systems, not only on PC but also for mobile phones like Android and Ios.

Tentative Submission: Please go here: https://github.com/sprites20/Sprites-RAG-Project/tree/main The program integrates with the TogetherAI API for the Large Language Model (LLM) and utilizes LlamaIndex Cohere embeddings and Cohere rerank to find and process documents. It is capable of retrieving relevant documentation locally and generating a program sequence for execution. This sequence, when run, generates output text that the system uses for its responses. The key enhancement this program provides to RAG's capabilities is the addition of a full code interpreter. This interpreter significantly improves RAG's ability to process information and generate accurate results, particularly in the realm of numeric computations. It also allows RAG to execute functions as requested by the user, expanding its utility and versatility. By leveraging the TogetherAI API and Cohere technology, this program enhances RAG's functionality across a wide range of tasks. It enables RAG to provide more detailed and accurate information, as well as perform calculations and execute functions, making it a more powerful tool for users seeking information and computational assistance.

A cross-platform application for creating AI agents and AGI, by coding using node editor, and can support more than Python language by using bindings for C/C++ (Cython), Java (jnius), C# (DLLs), Golang bindings, and whatever bindings is available. For this event, I integrated Stability's SDXL, Cohere's Aya model, TogetherAI for the LLM, but can also be configured for other models and providers. To create a video subtitle generator using faster-whisper, translated by the Cohere Aya model. Hopefully, the node editor will be included to run code asynchronously like in Unreal Engine's node editor and ComfyUI.

A low-code AI creation platform, where we can teach you how to integrate and create AI to help you in daily life and even monetize your skills, provide computing power, sell information products, you can turn high-level descriptions into modular components, connect inputs and outputs through intuitive node connectors, web browser integrations, AI chatbot, API connections, code execution, and a 3D renderer. Control your device's output hardware, play audio, display images, and more. Send nodes to the cloud provider of your choice, whether Colab, Replicate, Vercel, GCP, Azure, AWS, as long as it can run Python, will run functions in the cloud via a secure and encrypted Ngrok HTTPs server. With a marketplace for the AI, users may be able to add a paywall to the nodes and integrations they create, or even add ad integrations. You can even sell 3D models or make them move by the game creation platform. As of July 4, this is a work in progress.