linkun_215403

Is the `IterableDataset` the reason why, when using`tfrecord.torch.dataset.MultiTFRecordDataset` with `torch_xla.distributed.parallel_loader.ParallelLoader` gives the following type error- ``` Exception in device=TPU:0: object of type 'MultiTFRecordDataset' has no len() Traceback (most recent call...

@DelightRun awesome, but how can you define a `__len__` function when you are streaming from multiple TFRecords, using `tfrecord.torch.dataset.MultiTFRecordDataset` ?

I tried to implement `newMultiTFRecordDataset` which inherits from `torch.utils.data.Dataset` as follows: ``` class newMultiTFRecordDataset(): def __init__(self, data_pattern: str, index_pattern: typing.Union[str, None], splits: typing.Dict[str, float], description: typing.Union[typing.List[str], typing.Dict[str, str], None] =...

@DelightRun You just implemented the `len` function which was required. Nice! Thanks. But Can you just gimme the reason why my code dosenot works?

@DelightRun Moreover I get the same error after adding a `__len__()` function to `MultiTFRecordDataset`.

I would like to solve the issue.