Start with the installation instructions in install.ipynb.
Feel free to use the provided Dockerfile.
The API token which is to be set in the created .env (has to be renamed or copied from .env_template) file will be given during the actual hackathon session.
Once the installation is finished, you can start interacting with the remote open source llm using cutom_llm_interaction.ipynb.
The task will be to generate a text summary of a forecasting report providing highlights and general insights.
The report (see data/forecast_report.csv) will be of the form :
| business_unit | product_family | forecast_previous_month_production_2024 | forecast_current_month_production_2024 | forecast_current_month_marketing_2024 | possible_supply_for_2024 | |
|---|---|---|---|---|---|---|
| 0 | AL | Prozil | 103 | 104 | 120 | 100 |
| 1 | AL | Focix | 789 | 631 | 770 | 1000 |
| 2 | AL | Zentra | 42 | 34 | 35 | 50 |
| 3 | AL | Optix | 957 | 766 | 800 | 500 |
| 4 | AL | Mentix | 221 | 177 | 200 | 500 |
Feel free to fork and share your hacking results!