feat: 修改gradio需要的模型、依赖库、语言

Dockerfile

+FROM image.sourcefind.cn:5000/gpu/admin/base/jupyterlab-pytorch:2.2.0-python3.10-cuda12.1-ubuntu22.04 as base
+ARG IMAGE=voice-clone
+ARG IMAGE_UPPER=voice-clone
+ARG BRANCH=gpu
+RUN cd /root && git clone -b $BRANCH http://developer.hpccube.com/codes/chenpangpang/$IMAGE.git
+WORKDIR /root/$IMAGE/$IMAGE_UPPER
+RUN pip install -r requirements.txt
+
+
+#########
+# Prod  #
+#########
+FROM image.sourcefind.cn:5000/gpu/admin/base/jupyterlab-pytorch:2.2.0-python3.10-cuda12.1-ubuntu22.04
+ARG IMAGE=voice-clone
+ARG IMAGE_UPPER=voice-clone
+COPY chenyh/$IMAGE/frpc_linux_amd64_v0.2 /opt/conda/lib/python3.10/site-packages/gradio/
+RUN chmod +x /opt/conda/lib/python3.10/site-packages/gradio/frpc_linux_amd64_v0.2
+COPY chenyh/$IMAGE/coqui/XTTS-v2 /root/.local/share/tts/tts_models--multilingual--multi-dataset--xtts_v2
+COPY --from=base /opt/conda/lib/python3.10/site-packages /opt/conda/lib/python3.10/site-packages
+COPY --from=base /root/$IMAGE/$IMAGE_UPPER /root/$IMAGE_UPPER
+COPY --from=base /root/$IMAGE/启动器.ipynb /root/$IMAGE/start.sh /root/
\ No newline at end of file

voice-clone/app.py

-import spaces
 import gradio as gr
 import torch
 from TTS.api import TTS
@@ -11,7 +10,6 @@ device = "cuda"
 tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2").to(device)
 
 
-@spaces.GPU(enable_queue=True)
 def clone(text, audio, language):
     tts.tts_to_file(text=text, speaker_wav=audio, language=language, file_path="./output.wav")
     return "./output.wav"
@@ -21,10 +19,12 @@ iface = gr.Interface(fn=clone,
                      inputs=[gr.Textbox(label='Text'),
                              gr.Audio(type='filepath', label='Voice reference audio file'),
                              gr.Dropdown(label='Language',
-                                         choices=['en - English', 'fr - French', 'ar - Arabic', 'es - Spanish',
-                                                  'zh - Chinese', 'ru - Russian', 'pt - Portuguese', 'it - Italian',
-                                                  'de - German', 'nl - Dutch'])],
-                     outputs=gr.Audio(type='filepath'),
+                                         # choices=['en - English', 'fr - French', 'ar - Arabic', 'es - Spanish',
+                                         #          'zh - Chinese', 'ru - Russian', 'pt - Portuguese', 'it - Italian',
+                                         #          'de - German', 'nl - Dutch'])],
+                                         choices=['en', 'es', 'fr', 'de', 'it', 'pt', 'pl', 'tr', 'ru', 'nl', 'cs',
+                                                  'ar', 'zh-cn', 'hu', 'ko', 'ja', 'hi'])],
+                     outputs=gr.Audio(type='filepath', label='Generated audio file'),
                      title='Voice Clone',
                      description="""
                      by [Tony Assi](https://www.tonyassi.com/)
@@ -43,4 +43,4 @@ iface = gr.Interface(fn=clone,
                                 "./audio/Megan-Fox.mp3", "en - English"],
                                ["Hey there, it's me Jeff Goldblum. Type in whatever you'd like me to say.",
                                 "./audio/Jeff-Goldblum.mp3", "en - English"], ])
-iface.launch()
+iface.launch(server_name='0.0.0.0', share=True)
\ No newline at end of file

voice-clone/requirements.txt

 TTS
-torch
+torch
\ No newline at end of file
+gradio==4.26.0
\ No newline at end of file