fix(evaluate): Evaluate should not error when display_table=True

[christeefy]

Obtained an error when display_table=True is provided for Evaluate.
See simple reproducible example below:

from dspy.evaluate.evaluate import Evaluate
from dspy.evaluate.metrics import answer_exact_match
from dspy.predict import Predict

def new_example(question, answer):
    """Helper function to create a new example."""
    return dspy.Example(
        question=question,
        answer=answer,
    ).with_inputs("question")

devset = [new_example("What is 1+1?", "2")]

program = Predict("question -> answer")

ev = Evaluate(
    devset=devset,
    metric=answer_exact_match,
    display_table=True,
)
ev(program)  # TypeError: cannot do positional indexing on RangeIndex with these indexers [True] of type bool

This stems from isinstance(True, int) == True.

This PR contains a proposed fix with updated unit tests. Feel free to re-structure the unit tests as you see fit.

[arnavsinghvi11]

Thanks @christeefy ! great find!