[Enhancement] Retrieve multiple records at once with pipeline

Currently there is implementation details for saving | deleting multiple records at once using redis pipelines.
Example:

@py_test_mark_asyncio
async def test_delete_many(m):
    member1 = m.Member(
        id=0,
        first_name="Andrew",
        last_name="Brookins",
        email="a@example.com",
        join_date=today,
        age=38,
        bio="This is the user bio.",
    )
    member2 = m.Member(
        id=1,
        first_name="Kim",
        last_name="Brookins",
        email="k@example.com",
        join_date=today,
        age=34,
        bio="This is the bio for Kim.",
    )
    members = [member1, member2]
    result = await m.Member.add(members)
    assert result == [member1, member2]
    result = await m.Member.delete_many(members)
    assert result == 2
    with pytest.raises(NotFoundError):
        await m.Member.get(pk=member1.key())

I want to be able to get multiple records at once, so as to minimize network overhead.
Please make it possible

eg. query_members=[pk1, pk2]

result = await m.Member.get_many(query_members)

So something like the delete_many function:

    @classmethod
    async def get_many(
        cls,
        models: Sequence["RedisModel"],
        pipeline: Optional[redis.client.Pipeline] = None,
    ) -> int:
        db = cls._get_db(pipeline)

        for chunk in ichunked(models, 100):
            pks = [model.key() for model in chunk]
            await cls._get(db, *pks)

        return len(models)

Start by locating the existing delete_many implementation and the model retrieval entry points it uses. Define get_many around the requested primary keys and pipeline behavior described in the issue, then add coverage confirming multiple records are retrieved with reduced network overhead and the expected result shape.