tsurumeso

There is no need to split the audio in advance. This is because it is automatically split to the appropriate size during training. if you don't have enough VRAM, it...

Thank you for your kind words. You can use the `--pretrained_model` option like this: ```sh python train.py --dataset path/to/dataset --pretrained_model path/to/pretrained/model --reduction_rate 0.5 --mixup_rate 0.5 --gpu 0 ```

It is normal because the loss function is changed from v6. Also `-` indicates a negative value.

Of course, the smaller the loss, the better. 0 is better than 0.1 and -0.1 is better than 0.