HearSay makes staying on top of your favorite sports teams and players easy with user profiles and fluid question and answering. See our live demo to get a feel for what HearSay can do!
To use and tinker with HearSay locally, follow these steps:
- Clone this repository
- Determine whether you want to use our fine-tuned sentiment-analyzation & entity-extraction models, or if you'd rather use the base models via the HuggingFace API.
- If fine-tuned, reach out to bera@umich.edu for access to the files (we don't host them in this repo). Then, in your local project directory create an environment variable using:
export FINE_TUNED=1If API, create the following environment variables instead:
export FINE_TUNED=0
export HUGGINGFACE_API_KEY="api key you create using link above"- In either case, you need to create a serp api key and create an environment variable using
export SERP_API_KEY="api key"- With these prerequisites met, you can run HearSay using these commands in separate shell windows:
# back-end
python3 -m venv env
source env/bin/activate && pip3 install -r requirements.txt
python3 main.py// front-end
cd hearsay-client
npm install
npm start- Navigate to localhost:3000 and start using HearSay!