dvc-streamlit-example/st_scripts/st_list_model.py

import flatten_dict
import git
import pandas as pd
import streamlit as st

from st_scripts.st_utils import MODELS_PARAMETERS, MODELS_EVALUATION_METRICS, MODELS_COMMITS


def get_model_info(model_commit: git.Commit) -> dict:
    model_commit_info = {
        "hash": model_commit.hexsha,
        "message": model_commit.message,
        "committed_datetime": str(model_commit.committed_datetime),
        "committer": str(model_commit.committer),
    }

    return {
        "commit": model_commit_info,
        **MODELS_PARAMETERS[model_commit.hexsha],
        "evaluation": MODELS_EVALUATION_METRICS[model_commit.hexsha],
    }


MODELS_DF = (
    pd.DataFrame([
        flatten_dict.flatten(get_model_info(model_commit), reducer="dot")
        for model_commit in MODELS_COMMITS
    ])
    .sort_values("commit.committed_datetime").reset_index(drop=True)
    .sort_values("commit.committed_datetime", ascending=False)
)


def st_list_model():
    st.markdown("### Models list")

    columns_checkboxes_beta_columns = st.columns(5)

    columns_checkboxes_beta_columns[0].text("Show columns for")
    show_commit_columns = columns_checkboxes_beta_columns[1].checkbox("commits", value=True)
    show_model_columns = columns_checkboxes_beta_columns[2].checkbox("model parameters", value=True)
    show_train_columns = columns_checkboxes_beta_columns[3].checkbox("train parameters", value=True)
    show_evaluation_columns = columns_checkboxes_beta_columns[4].checkbox("evaluation metrics", value=True)

    #%%
    def show_column(column: str) -> bool:
        return (
            (show_commit_columns or not column.startswith("commit."))
            and (show_model_columns or not column.startswith("model."))
            and (show_train_columns or not column.startswith("train."))
            and (show_evaluation_columns or not column.startswith("evaluation."))
        )

    st.table(MODELS_DF.filter(items=[column for column in MODELS_DF if show_column(column)]))