Closed trigsoft closed 3 weeks ago
@trigsoft Can you provide your code to help me reproduce this problem
https://colab.research.google.com/drive/1PHd9Vy6Mf5hNXcudNl8nV9iDv4ht2LTq?usp=sharing
I will send a mail to you with a ref audio. I cannot gen an proper audio with the same content....
@trigsoft you need to clone bigvgan follow README instruction, or the following command
More details is written on the README
Can you send me the link please? I don't see the git clone line?
This is what I see
Same error.
!git clone https://github.com/SWivid/F5-TTS.git %cd F5-TTS/src/third_party !git clone https://github.com/NVIDIA/BigVGAN.git %cd ../.. !pip install -e .
Same error.
!git clone https://github.com/SWivid/F5-TTS.git %cd F5-TTS/src/third_party !git clone https://github.com/NVIDIA/BigVGAN.git %cd ../.. !pip install -e .
have you changed the bigvgan source code?
Same error. !git clone https://github.com/SWivid/F5-TTS.git %cd F5-TTS/src/third_party !git clone https://github.com/NVIDIA/BigVGAN.git %cd ../.. !pip install -e .
have you changed the bigvgan source code?
Current code: !pip install torch==2.3.0+cu118 torchaudio==2.3.0+cu118 --extra-index-url https://download.pytorch.org/whl/cu118 !git clone https://github.com/SWivid/F5-TTS.git %cd F5-TTS/src/third_party !git clone https://github.com/NVIDIA/BigVGAN.git %cd ../.. !pip install -e .
Modified code:
import os import sys sys.path.append(os.path.dirname(os.path.abspath(file)))
import os import json from pathlib import Path from typing import Optional, Union, Dict
import torch import torch.nn as nn from torch.nn import Conv1d, ConvTranspose1d from torch.nn.utils import weight_norm, remove_weight_norm
import activations from utils import init_weights, get_padding from alias_free_activation.torch.act import Activation1d as TorchActivation1d from env import AttrDict
will close this issue, feel free to open if further questions
I tried on colab. it works fine with default vocoder.
It returned the error when using bigvgan.