search

tsurumeso / vocal-remover

Vocal Remover using Deep Neural Networks
MIT License
1.47k stars 215 forks source link

ZeroDivisionError #90

Closed Z2r-YT closed 2 years ago

Z2r-YT commented 2 years ago

tried using the solution one has already posted here, nothing fixed it. anyone help?

python train.py --dataset C:\Users\blank\Desktop\vocal-remover-v4.0.0\Training --reduction_rate 5 --mixup_rate 5 --gpu 0 

0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
# epoch 0
  * inner epoch 0
Traceback (most recent call last):
  File "train.py", line 224, in <module>
    main()
  File "train.py", line 195, in main
    train_loss = train_inner_epoch(
  File "train.py", line 48, in train_inner_epoch
    return sum_loss / len(X)
ZeroDivisionError: division by zero
tsurumeso commented 2 years ago

Is the dataset correctly placed and are there at least a few tens of pairs? By the way, --reduction_rate and --mixup_rate should be more than 0 and less than 1.