Add option for loading audio with video

### Describe the bug

Currently, `torchcodec` don't allow extracting `Audio` from `Video` https://github.com/meta-pytorch/torchcodec/issues/1158, but when I upload videos with audio to hub using `videofolder`, then this is not possible to retrieve audio from it. Probably `VideoDecoder` can be extended with `audio` parameter to retrieve this information

### Steps to reproduce the bug

```python
from datasets import load_dataset

test_ds = load_dataset("videofolder", data_dir="/path/to/video")
# uploaded version
# test_ds = load_dataset("Samoed/testds")

test_ds["train"][1]["video"]
#