A series of playgrounds and notebooks to utilize tool-calling. Feel free to contribute tools to the repository.

CLI for automating the creation of subtitles using whisper and pyannote. Accepts batch or single audio and video files.

Segmind Vega realtime demo using streamlit. Images load on keyup.

FastAPI + Streamlit interface for OpenAI Whisper-large-v3 with youtube-to-mp3

Simple interface to upload, transcribe, diarize, and create subtitles using whisper-large-v3, pyannote/segmentation 3.0 and pyannote/speaker-diarization-3.1 (CLI in development)

Voice Generation in 35 languages and translation for over 100 languages.

This repo contains a gradio application with an interface for latent-consistency/lcm-lora-sdxl. The interface includes the LCM lora as well as the papercut lora from TheLastBen/Papercut_SDXL

Kosmos-2-Patch14-224: Multi-modal Object Detection

This repo contains a code tutorial using the owlv2-base-patch16-ensemble model from Google for multi-modal object detection

miniAGI is a Streamlit application that leverages the zero-shot ReAct Agent from Langchain with vectorsearch and many various integrations.

This repo contains an api and demo application for the zephyr-7b-beta model from HuggingFace. The base model is mistral-7b-v0.1 and this is the second zephyr finetune from the huggingface team.

This repo contains an api with postgreSQL database and a demo application for the zephyr-7b-alpha model from Huggingface.

This repo contains a simple UI for the segmind/SSD-1B model. This model is a lightweight implementation of SDXL. It should consume roughly 17GB of VRAM with both SSD-1B and SDXL 1.0 simultaneously.

This project is an API with a demo application utilizing facebook's detr-resnet-50 model for object detection in both image and video. The usage in video is experimental as the original model is tr…

This repo contains a tutorial for a finetune of mistralai/mistral-7b-v0.1. More information can be found in their paper: https://huggingface.co/papers/2310.06825

This repo contains the code and data used to finetune microsoft's phi 1_5 model. The finetune contains roughly 1M additional tokens. The datasets used are a combination of 200k Stanford Alpaca exam…

This repo is a small RESTful API for inference of the orca finetune of the Mistral-7B LLM

RESTful API for Micrososoft's phi 1.5 billion paramater model. The model utilizes a "textbooks are all you need approach." More information can be found in their paper: https://arxiv.org/abs/2309.0…

This repo is an extension of the phi_1.5_api but includes a chat application demo of the endpoints. Available utilizing the research license provided by Microsoft.

GameGPT is a langchain application that allows the user to configure a gymnasium environment to train GPT to play games and evaluate the logic used.

GPT-Debate is a simulation where each agent has access to tools and a predefined debate topic. They will engage in a debate with access to the internet so they may use current information.

This is a repo of plugins available for miniAGI. miniAGI and its documentation are available at http://www.github.com/tdolan21/miniAGI

This is my profile readme for dynamic content