找不到MOSS-Audio-Tokenizer的config.json文件

[DHfusion]

我将Basic Usage示例代码里的

processor = AutoProcessor.from_pretrained( pretrained_model_name_or_path, trust_remote_code=True, )
里面的 pretrained_model_name_or_path替换为下载到本地的MOSS-TTS-Local-Transformer文件夹路径后，运行代码抛出error：

'[WinError 10060] 由于连接方在一段时间后没有正确答复或连接的主机没有反应，连接尝试失败。' thrown while requesting HEAD https://huggingface.co/OpenMOSS-Team/MOSS-Audio-Tokenizer/resolve/main/config.json

我又下载了MOSS-Audio-Tokenizer项目里的文件，把 pretrained_model_name_or_path替换为MOSS-Audio-Tokenizer，但又出现error：
ValueError: Unrecognized processing class in D:\MOSS-Audio-Tokenizer. Can't instantiate a processor, a tokenizer, an image processor or a feature extractor for this model. Make sure the repository contains the files of at least one of those processing classes.

麻烦您有空看一看，这是什么原因呢

[alpacaking]

First, please make sure you have downloaded the latest Hugging Face weights for OpenMOSS-Team/MOSS-TTS-Local-Transformer.
Then, you can try:

pretrained_model_name_or_path = "/path/to/MOSS-TTS-Local-Transformer"
processor = AutoProcessor.from_pretrained(
    pretrained_model_name_or_path,
    trust_remote_code=True,
    codec_path="/path/to/MOSS-Audio-Tokenizer"
)

[DHfusion]

First, please make sure you have downloaded the latest Hugging Face weights for OpenMOSS-Team/MOSS-TTS-Local-Transformer. Then, you can try:

pretrained_model_name_or_path = "/path/to/MOSS-TTS-Local-Transformer"
processor = AutoProcessor.from_pretrained(
pretrained_model_name_or_path,
trust_remote_code=True,
codec_path="/path/to/MOSS-Audio-Tokenizer"
)

thank you. it works. However, when I ran this Basic Usage code, I encountered anthor error:

OSError: Could not load this library: D:\miniconda_env_3_mosstts\Lib\site-packages\torchcodec\libtorchcodec_core4.dll
[end of libtorchcodec loading traceback].

this is my package version:
torch 2.9.1
torchaudio 2.9.1
torchcodec 0.10.0

[WangJinbaekhyun]

The error is caused by a version mismatch between torchcodec and your PyTorch installation.

You can fix this by downgrading torchcodec to a compatible version:

pip install torchcodec==0.8.1