search

2noise / ChatTTS

A generative speech model for daily dialogue.
https://2noise.com
GNU Affero General Public License v3.0
29.68k stars 3.24k forks source link

Load model error加载模型未初始化问题 #78

Closed yestolife closed 2 months ago

yestolife commented 2 months ago

An error about loading model. When I run chat.load_models()， return the info as:

INFO:ChatTTS.core:use cuda:0
WARNING:ChatTTS.core:vocos not initialized.
WARNING:ChatTTS.core:gpt not initialized.
WARNING:ChatTTS.core:tokenizer not initialized.
WARNING:ChatTTS.core:dvae not initialized.
yestolife commented 2 months ago

问题已解决，引起不能初始化各模型的原因是，用pip install ChatTTS的版本和这里的版本不同，不需要pip install，直接在本工程文件文件夹中运行infer.ipynb。

另外可以从huggingface的镜像网站下载模型文件，从本机加载模型文件要修改一点代码：

chat = ChatTTS.Chat()
chat.load_models(source = 'local', local_path='/The Path of Model')
zhijianglee commented 2 months ago

load models doesn't even accept source and local path as parameter

zhijianglee commented 2 months ago

This is not working for me

zhchx601 commented 2 months ago

已解决。不要pip install，直接从GitHub clone即可。 Do not use pip install, use clone instead.

zgimszhd61 commented 2 months ago

Traceback (most recent call last): File "/Users/a0000/mywork/commonLLM/opensource/nnnew/chatTTSbbb/ChatTTS/xchatx.py", line 16, in chat.load_models(model_dir) File "/Users/a0000/mywork/commonLLM/opensource/nnnew/chatTTSbbb/ChatTTS/ChatTTS/core.py", line 61, in load_models self._load({k: os.path.join(download_path, v) for k, v in OmegaConf.load(os.path.join(download_path, 'config', 'path.yaml')).items()}, kwargs) ^^^^^^^^^^^^^ UnboundLocalError: cannot access local variable 'download_path' where it is not associated with a value

wudi-7mi commented 2 months ago

Traceback (most recent call last): File "/Users/a0000/mywork/commonLLM/opensource/nnnew/chatTTSbbb/ChatTTS/xchatx.py", line 16, in chat.load_models(model_dir) File "/Users/a0000/mywork/commonLLM/opensource/nnnew/chatTTSbbb/ChatTTS/ChatTTS/core.py", line 61, in load_models self._load({k: os.path.join(download_path, v) for k, v in OmegaConf.load(os.path.join(download_path, 'config', 'path.yaml')).items()}, kwargs) ^^^^^^^^^^^^^ UnboundLocalError: cannot access local variable 'download_path' where it is not associated with a value

试试看把这句话： chat.load_models(source='locals', local_path='YOUR LOCAL PATH') 中的locals改成local

作者快改改这个大坑......

qdrk commented 2 months ago

+1

chief11712138 commented 2 months ago

Not working for me. load_models does not accept parameter 'source' and 'local_path'

StarCompute commented 2 months ago

问题已解决，引起不能初始化各模型的原因是，用pip install ChatTTS的版本和这里的版本不同，不需要pip install，直接在本工程文件文件夹中运行infer.ipynb。

另外可以从huggingface的镜像网站下载模型文件，从本机加载模型文件要修改一点代码：

chat = ChatTTS.Chat()
chat.load_models(source = 'local', local_path='/The Path of Model')

这个问题解决了，其他问题还是一大堆

shawzhb commented 2 months ago

TypeError: Chat.load_models() got an unexpected keyword argument 'source'. how to solve this?