Miso
Developed For
UofTHacks X
Tech Stack
React.js, Node.js, Python, Cohere, FastAPI, Uvicorn, Typescript, Figma
Roles
Developer, Designer
Miso is a large language model that is trained on 23000 tweets, messages, and captions to identify whether chat messages are gender discriminatory or toxic.
Won Best Hack to Promote Women in STEM at UoftHacks X
The API developed with Python utilizes Cohere's Natural Language Processing to create an AI model. A variety of research datasets have been consolidated to train the model on the Cohere platform.
The backend utilizes a REST API deployed with FastAPI and Uvicorn and can be fetched by chat bots on Discord, Teams, Slack, or any other web based chat service
Our Discord Bot demo is created in Node.js using the Discord.js library and the Web Demo utilizes ReactJS and Typescript to display the LLM in a web format.
To create an intentional and intuitive combination of UI & UX, product mockups were crafted using Figma.
Explore the web app, or check out the GitHub repository and Figma designs through the links below.
