用custom方式load在huggingface上下载的模型时报错哈希值不匹配
报错信息是：
```
D:\ChatTTS\asset\gpt\config.json sha256 hash mismatch.
check models in custom path D:\ChatTTS failed.
vocos not initialized.
gpt not initialized.
tokenizer not initialized.
embed not initialized.
decoder not initialized.
Traceback (most recent call last):
File "D:\VoxMeeter\tts_service\tts_server.py", line 11, in <module>
wavs = chat.infer(texts)
^^^^^^^^^^^^^^^^^
File "D:\anaconda\Lib\site-packages\ChatTTS\core.py", line 263, in infer
for wavs in res_gen:
File "D:\anaconda\Lib\site-packages\ChatTTS\core.py", line 402, in _infer
assert self.has_loaded(use_decoder=use_decoder)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AssertionError
```
运行脚本是：
```
import ChatTTS
import torch
import torchaudio
model_path = r"D:\ChatTTS"
chat = ChatTTS.Chat()
chat.load(compile=False,source="custom",custom_path=model_path) # Set to True for better performance
texts = ["测试1", "测试2"]
wavs = chat.infer(texts)
for i in range(len(wavs)):
"""
In some versions of torchaudio, the first line works but in other versions, so does the second line.
"""
try:
torchaudio.save(f"basic_output{i}.wav", torch.from_numpy(wavs[i]).unsqueeze(0), 24000)
except:
torchaudio.save(f"basic_output{i}.wav", torch.from_numpy(wavs[i]), 24000)
```
文件是从huggingface上直接git下来的，没动任何东西，因为用local模式load的时候一直显示下载模型失败，就自己下载了，但也load不了，不知道啥问题，检查了一下config.json好像没啥异常情况，不知道是不是版本的问题，仓库是从github克隆的最新版本
0 条评论