Text Eval End-to-End

import zeroeval as ze

ze.init()

dataset = ze.Dataset(
    "qa-text-e2e",
    data=[
        {"row_id": "1", "question": "Capital of France?", "answer": "Paris"},
        {"row_id": "2", "question": "Capital of Germany?", "answer": "Berlin"},
        {"row_id": "3", "question": "Capital of Japan?", "answer": "Tokyo"},
    ],
)
dataset.push()

@ze.task(outputs=["prediction"])
def answer(row):
    # Replace with provider call.
    return {"prediction": row.answer}

@ze.evaluation(mode="row", outputs=["exact_match"])
def exact_match(row, answer_col, prediction_col):
    return {"exact_match": int(answer_col == prediction_col)}

@ze.evaluation(mode="column", outputs=["accuracy"])
def accuracy(exact_match_col):
    total = len(exact_match_col)
    return {"accuracy": (sum(exact_match_col) / total) if total else 0.0}

run = dataset.eval(
    answer,
    execution=ze.ExecutionConfig(
        workers=8,
        timeout_s=30,
        retry=ze.RetryPolicy(max_attempts=3),
    ),
    checkpoint=ze.CheckpointConfig(flush_every_rows=50, flush_every_seconds=5.0),
)

run = run.score(
    [exact_match, accuracy],
    column_map={
        "exact_match": {
            "answer_col": "answer",
            "prediction_col": "prediction",
        },
        "accuracy": {"exact_match_col": "exact_match"},
    },
)

print("eval_id:", run.id)
print("metrics:", run.metrics)
print("health:", run.health)

@ze.evaluation(mode="run", outputs=["accuracy_mean", "accuracy_n"])
def accuracy_over_runs(all_runs):
    vals = [r.metrics["accuracy"] for r in all_runs if "accuracy" in r.metrics]
    return {
        "accuracy_mean": (sum(vals) / len(vals)) if vals else 0.0,
        "accuracy_n": len(vals),
    }

all_runs = run.repeat(5).to_list()
aggregate = all_runs[0]
aggregate.run_metrics([accuracy_over_runs], all_runs=all_runs)
print(aggregate.metrics)