feat: exploration - analysis of chat history and streamlit dashboard #10

	# This workflow will install Python dependencies and run AI regression tests with a single version of Python
	# For more information see: https://docs.github.com/en/actions/automating-builds-and-tests/building-and-testing-python

	name: LLM regression tests

	on:
	# weekdays at 6am
	schedule:
	- cron: '0 6 * * 1-5'
	# or on-demand
	workflow_dispatch:
	# or on push to main
	push:
	branches:
	- 'main'

	permissions:
	contents: read

	jobs:
	ai-tests:

	runs-on: ubuntu-latest

	steps:
	- name: Checkout code
	uses: actions/checkout@v3

	- name: Set up Python 3.11
	uses: actions/setup-python@v3
	with:
	python-version: "3.11"

	- name: Set up Poetry
	uses: abatilo/actions-poetry@v2
	with:
	poetry-version: 1.7.0

	- name: Build containers
	run: \|
	mkdir -p data/elastic/
	chmod 777 data/elastic/

	cp .env.test .env
	echo AZURE_OPENAI_ENDPOINT=${{ secrets.AZURE_OPENAI_ENDPOINT }} >> .env
	echo AZURE_OPENAI_API_KEY=${{ secrets.AZURE_OPENAI_API_KEY }} >> .env
	echo AZURE_OPENAI_MODEL=azure/gpt-4 >> .env
	echo OPENAI_API_VERSION=2024-02-01 >> .env
	sed -i 's/^EMBEDDING_MODEL=.*/EMBEDDING_MODEL=all-mpnet-base-v2/' .env

	docker compose up -d --wait elasticsearch
	poetry install --no-root --no-ansi --with dev,ai,api --without worker
	poetry run download-model --embedding_model all-mpnet-base-v2

	- name: Run DeepEval Unit Tests
	run: poetry run deepeval test run core_api/tests/test_ai.py --mark ai

Provide feedback