[EmoGen] run "bash Piano_train.sh" got "The maximum length exceeds 1280. Please resample the dataset"

[lajsean]

Followed its instruction on readMe.md file, when running bash Piano_train.sh on EmoGen, got following error. Anyone has any idea? Thanks in advance!

AssertionError: Caught AssertionError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/home/user1/Dev/ai-musicGen/muzic/emogen/venv-3.8.1/lib/python3.8/site-packages/torch/utils/data/_utils/worker.py", line 308, in _worker_loop
data = fetcher.fetch(index)
File "/home/user1/Dev/ai-musicGen/muzic/emogen/venv-3.8.1/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py", line 51, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/user1/Dev/ai-musicGen/muzic/emogen/venv-3.8.1/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py", line 51, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/user1/Dev/ai-musicGen/muzic/emogen/linear_decoder/controlled_task.py", line 44, in getitem
assert len(sample["source"]) <= self.args.truncated_length + 2, f"The maximum length exceeds {self.args.truncated_length}. Please resample the dataset."
AssertionError: The maximum length exceeds 1280. Please resample the dataset.

[lajsean]

finally worked around it by using default value 8192 to replace 1280