model-eval-be / README.md
ecemumutlu's picture
Update readme
121d926
metadata
title: Model Eval Be
emoji: πŸ†
colorFrom: green
colorTo: green
sdk: docker
pinned: false

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Setup Environment

python3 -m env model_be
source model_be/bin/activate
pip install -r requirements.txt

Secret Variables

  • FastAPI uses "AUTH_UUID" as an environment variable to authenticate. Add this variable to your .env file on development mode (local deployment).
  • If Huggingface space is used, set AUTH_UUID as a secret. (Set the same secret value in metunlp/backend)

Add New Task (Development)

Under virtual environment lib/python3/site-packages/lm_eval/tasks , add your new task. Once you are certain that the task works, add the task to the github repository (which is specified in requirements.txt file)

Run Application (port: 8080)

Change port or host on app.py if needed.

python app.py