vits2

vits2 copied to clipboard

DistributedBucketSampler division by zero problem

1

In the `DistributedBucketSampler ` class, it's possible that the first bucket has no elements and thus results in a `Division by zero` error. ```python class DistributedBucketSampler(torch.utils.data.distributed.DistributedSampler): # ... def _create_buckets(self):...

y1guo

Hi, does there any Chinese support?

12

Hi, does there any Chinese support?

lucasjinreal

AssertionError: min value is nan

5

Traceback (most recent call last): File "/home/work/miniconda3/envs/vits2/lib/python3.11/site-packages/torch/multiprocessing/spawn.py", line 69, in _wrap fn(i, *args) File "/home/work/语音相关/文字转语音/vits2/train_ms.py", line 92, in run train_and_evaluate(rank, epoch, hps, [net_g, net_d], [optim_g, optim_d], [scheduler_g, scheduler_d], scaler, [train_loader,...

wlz987

Hey, thanks for the implementation. Are you guys also planning to release the pre-trained weights anytime soon?

chiragoel

When I train with data from two custom speakers, why is the resulting audio unclear, and why is there an electric noise?[](url)

pengpengtao