lenML / Speech-AI-Forge

🍦 Speech-AI-Forge is a project developed around TTS generation model, implementing an API Server and a Gradio-based WebUI.
https://huggingface.co/spaces/lenML/ChatTTS-Forge
GNU Affero General Public License v3.0
711 stars 87 forks source link

本地部署报错,无法生成,貌似无法使用cuda导致 #1

Closed alex9441 closed 4 months ago

alex9441 commented 4 months ago

前端可以启动,点击生成报‘Error’ 日志输出 RuntimeError: Attempting to deserialize object on a CUDA device but torch.cuda.is_available() is False. If you are running on a CPU-only machine, please use torch.load with map_location=torch.device('cpu') to map your storages to the CPU. DEBUG:matplotlib.pyplot:Loaded backend agg version v2.2. ERROR:modules.ssml:apply style failed, Attempting to deserialize object on a CUDA device but torch.cuda.is_available() is False. If you are running on a CPU-only machine, please use torch.load with map_location=torch.device('cpu') to map your storages to the CPU. INFO:modules.ssml:collect len(segments): 1 DEBUG:matplotlib.pyplot:Loaded backend QtAgg version 5.15.10. 看起来是无法使用cuda加速导致的。我计算机应该没问题,另外两个同样基于ChatTTS的webUI都可以运行。

handles98 commented 4 months ago

我这里可以正常生成,是不是没安装GPU版的torch?安装试下,cu118后面换成你系统的CUDA版本号 pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

alex9441 commented 4 months ago

从新CONDA了虚拟环境,程序莫名其妙的跑起来了。。。。 好吧。 谨祝作者码运昌隆!