[ISSUE] 语音复刻出现问题，提示：Could not initialize NNPACK! Reason: Unsupported hardware

[No-22-Github]

确认清单

• [X] 我已经阅读过 README.md 和 dependencies.md 文件
• [X] 我已经确认之前没有 issue 或 discussion 涉及此 BUG
• [X] 我已经确认问题发生在最新代码或稳定版本中
• [X] 我已经确认问题与 API 无关
• [X] 我已经确认问题与 WebUI 无关
• [X] 我已经确认问题与 Finetune 无关

你的issues

部署完成后tts正常，语音复刻会报错：

python webui.py
2024-10-02 15:18:01,000 - modules.core.models.tts.CosyVoiceModel - INFO - No CosyVoice model found
2024-10-02 15:18:02,351 - httpx - INFO - HTTP Request: GET https://api.gradio.app/gradio-messaging/en "HTTP/1.1 200 OK"
2024-10-02 15:18:03,132 - modules.devices.devices - INFO - Using half precision: torch.float16
2024-10-02 15:18:03,133 - modules.devices.devices - INFO - Using device: cuda
2024-10-02 15:18:03,390 - modules.webui.app - INFO - WebUI module initialized
2024-10-02 15:18:03,391 - modules.webui.localization - INFO - Loaded localization file /home/no22/ChatTTS-Forge/language/zh-CN.json
2024-10-02 15:18:04,576 - httpx - INFO - HTTP Request: GET https://checkip.amazonaws.com/ "HTTP/1.1 200 "
Running on local URL:  http://0.0.0.0:7860
2024-10-02 15:18:04,671 - httpx - INFO - HTTP Request: GET https://api.gradio.app/pkg-version "HTTP/1.1 200 OK"
2024-10-02 15:18:04,714 - httpx - INFO - HTTP Request: GET http://localhost:7860/startup-events "HTTP/1.1 200 OK"
2024-10-02 15:18:04,798 - httpx - INFO - HTTP Request: HEAD http://localhost:7860/ "HTTP/1.1 200 OK"

To create a public link, set `share=True` in `launch()`.
2024-10-02 15:18:55,651 - modules.core.models.zoo.ChatTTS - INFO - Loading ChatTTS models
2024-10-02 15:18:55,689 - modules.repos_static.ChatTTS.ChatTTS.core - INFO - try to load from local: ./models/ChatTTS
2024-10-02 15:18:55,689 - modules.repos_static.ChatTTS.ChatTTS.core - INFO - checking assets...
2024-10-02 15:18:59,418 - modules.repos_static.ChatTTS.ChatTTS.core - INFO - all assets are already latest.
2024-10-02 15:18:59,862 - modules.repos_static.ChatTTS.ChatTTS.core - INFO - vocos loaded.
2024-10-02 15:19:00,278 - modules.repos_static.ChatTTS.ChatTTS.core - INFO - dvae loaded.
2024-10-02 15:19:08,873 - modules.repos_static.ChatTTS.ChatTTS.core - INFO - gpt loaded.
2024-10-02 15:19:09,425 - modules.repos_static.ChatTTS.ChatTTS.core - INFO - decoder loaded.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
2024-10-02 15:19:09,440 - modules.repos_static.ChatTTS.ChatTTS.core - INFO - tokenizer loaded.
2024-10-02 15:19:09,493 - modules.core.models.zoo.ChatTTS - INFO - ChatTTS models loaded
2024-10-02 15:19:13,306 - modules.Enhancer.ResembleEnhance - INFO - Loading ResembleEnhance model
2024-10-02 15:19:14,791 - modules.repos_static.resemble_enhance.enhancer.lcfm.lcfm - INFO - Freeze cfm
/home/no22/ChatTTS-Forge/modules/Enhancer/ResembleEnhance.py:34: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  state_dict = torch.load(
2024-10-02 15:19:18,678 - modules.Enhancer.ResembleEnhance - INFO - ResembleEnhance model loaded
[W1002 15:19:18.658460158 NNPACK.cpp:61] Could not initialize NNPACK! Reason: Unsupported hardware.

我的系统配置： 使用python 3.10+torch: 2.4.1+cu121+ffmpeg: n7.0.2 显卡驱动是最新的，nvidia-smi和nvcc-V正常，cpu支持avx指令集，不支持avx2