Got error when running training code.

[alexlfyang]

Training mel:   0%|          | 0/200000 [00:11<?, ?it/s]
Traceback (most recent call last):
  File "/home/ap/miniconda3/envs/ppg_train/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/home/ap/miniconda3/envs/ppg_train/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/home/ap/ppgs/ppgs/train/__main__.py", line 45, in <module>
    main(**vars(parse_args()))
  File "/home/ap/ppgs/ppgs/train/__main__.py", line 24, in main
    ppgs.train(dataset, directory, gpu)
  File "/home/ap/miniconda3/envs/ppg_train/lib/python3.10/contextlib.py", line 79, in inner
    return func(*args, **kwds)
  File "/home/ap/ppgs/ppgs/train/core.py", line 231, in train
    evaluate_fn(valid_loader, 'valid')
  File "/home/ap/ppgs/ppgs/train/core.py", line 348, in evaluate
    metrics.update(predicted_ppgs, indices)
  File "/home/ap/ppgs/ppgs/evaluate/metrics.py", line 39, in update
    metric.update(predicted_logits, target_indices)
  File "/home/ap/ppgs/ppgs/evaluate/metrics.py", line 175, in update
    jsd = ppgs.distance(
  File "/home/ap/ppgs/ppgs/core.py", line 437, in distance
    ppgX = torch.mm(distance.similarity_matrix.T ** exponent, ppgX).T
TypeError: unsupported operand type(s) for ** or pow(): 'Tensor' and 'NoneType'

Any insight how to fix it?

[CameronChurchwell]

I mean this would seem to imply that exponent is None here, which should only happen if you set pggs.SIMILARITY_EXPONENT=None, but there could be a bug causing this as well... Maybe check what value of SIMILARITY_EXPONENT is being set after importing ppgs?