| import pandas as pd | |
| from display.utils import AutoEvalColumn | |
| from leaderboard.read_evals import get_raw_eval_results, EvalResult | |
| def get_leaderboard_df( | |
| results_path: str, requests_path: str, cols: list, benchmark_cols: list | |
| ) -> tuple[list[EvalResult], pd.DataFrame]: | |
| """Creates a dataframe from all the individual experiment results""" | |
| raw_data = get_raw_eval_results(results_path, requests_path) | |
| all_data_json = [v.to_dict() for v in raw_data] | |
| df = pd.DataFrame.from_records(all_data_json) | |
| df = df.sort_values(by=[AutoEvalColumn.average.name], ascending=False) | |
| df = df[cols].round(decimals=2) | |
| return raw_data, df | |