batch does not use serializers

[ydjin0602]

Describe the bug
When I use a batch producer with the key_serializer and value_serializer keys, the key and value are not serialized in any way, and therefore I get a TypeError.
When sending a single message via producer.send(), the serialization works as it should.

Environment (please complete the following information):

• aiokafka version 0.8.0
• kafka-python version 2.0.2
• Kafka Broker version: I use docker image confluentinc/cp-kafka:latest on my pc

Reproducible example

def _serializer(value):
      return json.dumps(value).encode("utf-8")

def _key_serializer(value):
      return value.encode("utf-8")

async def send_many_messages(
        self,
        messages: list[BaseSchema],
        topic: str,
        key: str,
        headers: List[Tuple[str, bytes]],
    ) -> None:
        producer = AIOKafkaProducer(
            bootstrap_servers=[f"{CONFIG.KAFKA_HOST}:{CONFIG.KAFKA_PORT}"],
            enable_idempotence=True,
            key_serializer=self._key_serializer,
            value_serializer=self._serializer,
        )
        try:
            await producer.start()
            batch = producer.create_batch()
            for message in messages:
                metadata = batch.append(
                    key=key,
                    headers=headers,
                    value=message,
                    timestamp=None,
                )
                if metadata is None:
                    partitions = await producer.partitions_for(topic=topic)
                    partition = random.choice(tuple(partitions))
                    await producer.send_batch(
                        batch=batch,
                        topic=topic,
                        partition=partition,
                    )
                    logger.info(
                        f"{batch.record_count()} сообщений отправлено в "
                        f"Kafka топик {topic}!"
                    )
                    batch = producer.create_batch()
                    continue

            partitions = await producer.partitions_for(topic=topic)
            partition = random.choice(tuple(partitions))
            await producer.send_batch(
                batch=batch,
                topic=topic,
                partition=partition,
            )
            logger.info(
                f"{batch.record_count()})
        except (ValueError, KafkaConnectionError):
            raise KafkaConnectionError(
                message="some error"
            )
        finally:
            await producer.stop()

[ods]

Fixed in #887