Simulation Based Optimization in Python using Model-based Optimization Algorithms The Bayesian Optimization Algorithms run with botorch. Normally the newest version of botorch works, except of MorBO, here we need botorch 0.7.0.
Step 1: Install requirement.txt (ATTENTION! No Versions provided - will install the newest):
pip -r requirements.txt
Step 2: Create config.py file in the root directory and set the following variables:
SHEET_ID = "SET YOUR SHEET ID HERE" FIREBASE_CONFIG = "SET YOUR FIREBASE CONFIG PATH HERE" BUCKET = "SET YOUR BUCKET NAME HERE" BIGQUERY_DATASET = "SET YOUR BIGQUERY DATASET NAME HERE" GCLOUD_PROJECT = "SET YOUR GCLOUD PROJECT NAME HERE" GCLOUD_SERVICE_ACCOUNT = "SET YOUR GCLOUD SERVICE ACCOUNT PATH HERE" DB_NAME = "YOU DB NAME"
Step 3: initialize database with:
python backend/databases/init_sql.py
Step 4: Start experiments with:
python main.py