tfds.load downloads nsynth with missing parts

Short description
I am using tfds.load to download nsynth dataset. But no matter how many times I try, It's only downloaded up until nsynth-train.tfrecord-00975-of-01024. It seems like 976 to 1023 are missing.

Environment information

• Operating System: Windows 10
• Python version: 3.7.3
• tensorflow-datasets/tfds-nightly version: tensorflow-datasets@2.1.0
• tensorflow/tensorflow-gpu/tf-nightly/tf-nightly-gpu version: tensorflow@2.1.0

Reproduction instructions

datasets = tfds.load("nsynth", data_dir="data")

Link to logs

Dl Completed...: 100%|██████████████████████████████████████████████████████████| 989/989 [2:28:10<00:00,  1.89s/ file]

Dl Completed...: 100%|██████████████████████████████████████████████████████████| 989/989 [2:28:10<00:00,  8.99s/ file]

Dataset nsynth downloaded and prepared to data\nsynth\full\2.3.0. Subsequent calls will reuse this data.

Expected behavior
I expect the whole of the dataset to be downloaded.

Start by reproducing the issue with the shown tfds.load("nsynth", data_dir="data") call and compare the download log with the expected 1,024 shards. Investigate why the completed download reports only 989 files; done means the full NSynth dataset, including shards 976 through 1023, is downloaded and prepared.