Usage Guide
This guide provides comprehensive instructions for running the Main pipeline, from single experiment execution to large-scale batch processing. The template for running the Main pipeline is launcher/run_main_template.sh. You can make a copy of the template (e.g. launcher/run_main_ft.sh) and modify it to run the Main pipeline for your experiment set and data. Put the copy in the launcher/ directory. Usually you need to change the EXP_SET_NAME and the EXPERIMENTS array, which are marked with TODO in the bash script. Then call bash launcher/run_main_ft.sh to run with your data.
The runnable script (e.g. launcher/run_main_ft.sh) calls scripts/run_main.sh with the following arguments
EXP_SET_NAME: The name of the experiment setEXP_ID: The name of the experimentROOT_DIR: The root directory of your projectDATA_DIR: The directory of your dataCONFIG_DIR: The directory of your configOUT_DIR: The directory of your outputLOG_LEVEL: The log level
Then run_main.sh runs the src/main.py script to launch the Main pipeline.