fix

Dockerfile

@@ -17,7 +17,7 @@ RUN apt-get update && \
 # Copy only requirements.txt initially to leverage Docker cache
 WORKDIR /workspace
 COPY requirements.txt /workspace/
-RUN pip install --no-cache-dir -r requirements.txt
+RUN CMAKE_POLICY_VERSION_MINIMUM=3.5 pip install --no-cache-dir -r requirements.txt
 
 # Define a build-time argument for image type
 ARG IMAGE_TYPE=full
@@ -39,4 +39,4 @@ COPY . /workspace
 
 EXPOSE 9871 9872 9873 9874 9880
 
-CMD ["python", "api.py"]
+CMD ["python", "api.py"]

api.py

@@ -173,6 +173,9 @@ import config as global_config
 import logging
 import subprocess
 
+import nltk
+nltk.download('averaged_perceptron_tagger_eng')
+
 
 class DefaultRefer:
     def __init__(self, path, text, language):
@@ -1086,9 +1089,10 @@ async def tts_endpoint(request: Request):
 
 @app.get("/")
 async def tts_endpoint(
-        refer_wav_path: str = None,
-        prompt_text: str = None,
-        prompt_language: str = None,
+        refer_wav_path: str = "saotome/saotome-6s.wav",
+        prompt_text: str = "今日は友達と一緒に映画を見に行く予定ですが、天気が悪くて少し心配です。",
+        prompt_language: str = "all_ja",
+        character: str = "saotome",
         text: str = None,
         text_language: str = None,
         cut_punc: str = None,
@@ -1096,10 +1100,11 @@ async def tts_endpoint(
         top_p: float = 1.0,
         temperature: float = 1.0,
         speed: float = 1.0,
-        inp_refs: list = Query(default=[]),
+        inp_refs: list = Query(default=["saotome/saotome-10s.wav", "saotome/refs/ref1.wav", "saotome/refs/ref2.wav", "saotome/refs/ref3.wav", "saotome/refs/ref4.wav"]),
         sample_steps: int = 32,
         if_sr: bool = False
 ):
+    print(f"the base path is {refer_wav_path}")
     return handle(refer_wav_path, prompt_text, prompt_language, text, text_language, cut_punc, top_k, top_p, temperature, speed, inp_refs, sample_steps, if_sr)
 
 

requirements.txt

@@ -1,42 +1,40 @@
-numpy==1.23.4
-scipy
-tensorboard
-librosa==0.9.2
-numba==0.56.4
-pytorch-lightning
-gradio>=4.0,<=4.24.0
-ffmpeg-python
-onnxruntime; sys_platform == 'darwin'
-onnxruntime-gpu; sys_platform != 'darwin'
-tqdm
-funasr==1.0.27
-torch<2.4
-cn2an
-pypinyin
-pyopenjtalk>=0.3.4
-g2p_en
-torchaudio
-modelscope==1.10.0
-sentencepiece
-transformers>=4.43
-peft
-chardet
-PyYAML
-psutil
-jieba_fast
-jieba
-split-lang
-fast_langdetect>=0.3.0
-Faster_Whisper
-wordsegment
-rotary_embedding_torch
-ToJyutping 
-g2pk2
-ko_pron
-opencc; sys_platform != 'linux'
-opencc==1.1.1; sys_platform == 'linux'
-python_mecab_ko; sys_platform != 'win32'
-fastapi<0.112.2
-x_transformers
-torchmetrics<=1.5
-attrdict
+numpy==1.23.4  
+scipy==1.13.1  
+tensorboard==2.19.0  
+librosa==0.9.2  
+numba==0.56.4  
+pytorch-lightning==2.5.1  
+gradio==4.24.0  
+ffmpeg-python==0.2.0  
+onnxruntime==1.19.2 #(for macOS)  
+onnxruntime-gpu==1.19.2 #(for non-macOS)  
+tqdm==4.67.1  
+funasr==1.0.27  
+cn2an==0.5.23  
+pypinyin==0.53.0  
+pyopenjtalk==0.4.0  
+g2p-en==2.1.0  
+torchaudio==2.6.0  
+modelscope==1.10.0  
+sentencepiece==0.2.0  
+transformers==4.50.0  
+peft==0.15.0  
+chardet==5.2.0  
+PyYAML==6.0.2  
+psutil==7.0.0  
+jieba_fast==0.53  
+jieba==0.42.1  
+split-lang==2.1.0  
+fast-langdetect==0.3.1  
+faster-whisper==1.1.1  
+wordsegment==1.3.1  
+rotary-embedding-torch==0.8.6  
+ToJyutping==3.2.0  
+g2pk2==0.0.3  
+ko-pron==1.3  
+opencc==1.1.1 #(for Linux)  
+python-mecab-ko==1.3.7 #(not for Windows)  
+fastapi==0.112.1 # (since <0.112.2)  
+x-transformers==2.1.37  
+torchmetrics==1.5.0  
+attrdict==2.0.1