chichi-speech 0.1.0__py3-none-any.whl

chichi_speech/client.py

@@ -0,0 +1,41 @@
+
+import argparse
+import requests
+import sys
+import os
+
+def main():
+    parser = argparse.ArgumentParser(description="Client for Qwen3 TTS Service")
+    parser.add_argument("text", help="Text to synthesize")
+    parser.add_argument("-l", "--language", default="Chinese", help="Language code (default: Chinese)")
+    parser.add_argument("-o", "--output", default="output.wav", help="Output WAV file path")
+    parser.add_argument("--url", default="http://localhost:9090", help="Service URL (default: http://localhost:9090)")
+    
+    args = parser.parse_args()
+    
+    endpoint = f"{args.url.rstrip('/')}/synthesize"
+    payload = {
+        "text": args.text,
+        "language": args.language
+    }
+    
+    print(f"Sending request to {endpoint}...")
+    try:
+        response = requests.post(endpoint, json=payload, stream=True)
+        response.raise_for_status()
+        
+        with open(args.output, "wb") as f:
+            for chunk in response.iter_content(chunk_size=8192):
+                f.write(chunk)
+        
+        print(f"Success! Audio saved to: {os.path.abspath(args.output)}")
+        
+    except requests.exceptions.ConnectionError:
+        print(f"Error: Could not connect to service at {args.url}. Is the server running?")
+        sys.exit(1)
+    except Exception as e:
+        print(f"Error: {e}")
+        sys.exit(1)
+
+if __name__ == "__main__":
+    main()

chichi_speech/server.py

@@ -0,0 +1,117 @@
+import uvicorn
+from fastapi import FastAPI, HTTPException
+from fastapi.responses import StreamingResponse
+from pydantic import BaseModel
+import torch
+import soundfile as sf
+import io
+from qwen_tts import Qwen3TTSModel
+
+# Initialize FastAPI app
+app = FastAPI(title="ChiChi Speech Service")
+
+# Global variables
+model = None
+VOICE_PROMPT = None
+
+# Hardcoded reference constants for voice cloning
+# REF_AUDIO = "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen3-TTS-Repo/clone_2.wav"
+# REF_TEXT = "Okay. Yeah. I resent you. I love you. I respect you. But you know what? You blew it! And thanks to you."
+REF_AUDIO = "/Users/stevenliu/Downloads/0126_ori_voice.WAV"
+REF_TEXT = "你那日问我，我的眼里是如何看待我？元氏父兄我打你，他们试着训过，终点并非叛贼。你总说守护江阴义不容辞，守护江阴义不容辞，可我怕你守到这最后，你也住了那史书上殉国忠烈。"
+
+
+@app.on_event("startup")
+async def startup_event():
+    global model, VOICE_PROMPT
+    print("Loading Qwen3 TTS Model...")
+    # Initialize the model
+    # Using the same parameters as voice_clone_basic.py
+    model = Qwen3TTSModel.from_pretrained(
+        "Qwen/Qwen3-TTS-12Hz-1.7B-Base",
+        device_map="mps",
+        dtype=torch.float32,
+    )
+    
+    print("Creating Voice Clone Prompt...")
+    # Pre-compute the prompt using the hardcoded reference audio and text
+    # This corresponds to: prompt_items = tts.create_voice_clone_prompt(...)
+    # We default x_vector_only_mode to False as the variable 'xvec_only' 
+    # from the snippet is unknown, and typically such flags are optional.
+    VOICE_PROMPT = model.create_voice_clone_prompt(
+        ref_audio=REF_AUDIO,
+        ref_text=REF_TEXT,
+        # x_vector_only_mode=False
+    )
+    print("Service Ready.")
+
+class SynthesisRequest(BaseModel):
+    text: str
+    language: str = "Chinese"
+
+@app.post("/synthesize")
+async def synthesize(request: SynthesisRequest):
+    """
+    Synthesize speech using the pre-loaded voice clone prompt.
+    """
+    if not model or VOICE_PROMPT is None:
+        raise HTTPException(status_code=503, detail="Service not initialized")
+
+    try:
+        # Generate the voice clone
+        # corresponding to: return tts.generate_voice_clone(...)
+        wavs, sr = model.generate_voice_clone(
+            text=request.text,
+            language=request.language,
+            voice_clone_prompt=VOICE_PROMPT,
+        )
+        
+        # wavs[0] is the audio data. Ensure it's a numpy array for soundfile.
+        audio_data = wavs[0]
+        if hasattr(audio_data, "cpu"):
+             audio_data = audio_data.cpu().float().numpy()
+             
+        # Write to an in-memory buffer
+        buffer = io.BytesIO()
+        sf.write(buffer, audio_data, sr, format='WAV')
+        buffer.seek(0)
+        
+        return StreamingResponse(buffer, media_type="audio/wav")
+        
+    except Exception as e:
+        import traceback
+        traceback.print_exc()
+        raise HTTPException(status_code=500, detail=str(e))
+
+def main():
+    import argparse
+    import os
+    
+    parser = argparse.ArgumentParser(description="Qwen3 TTS Service")
+    parser.add_argument("--port", type=int, default=9090, help="Service port (default: 9090)")
+    parser.add_argument("--host", type=str, default="0.0.0.0", help="Service host (default: 0.0.0.0)")
+    parser.add_argument("--ref-audio", type=str, nargs="+", help="Path(s) to reference audio file(s) for voice cloning")
+    parser.add_argument("--ref-text", type=str, nargs="+", help="Reference text content(s) corresponding to the audio")
+
+    args = parser.parse_args()
+
+    # Environment variable overrides
+    if "PORT" in os.environ:
+        args.port = int(os.environ["PORT"])
+
+    # Update global configuration if arguments specific
+    global REF_AUDIO, REF_TEXT
+    if args.ref_audio:
+        REF_AUDIO = args.ref_audio
+    if args.ref_text:
+        REF_TEXT = args.ref_text
+
+    print(f"Starting server on {args.host}:{args.port}")
+    if args.ref_audio:
+        print(f"Overriding reference audio: {args.ref_audio}")
+
+    # Run the server
+    uvicorn.run(app, host=args.host, port=args.port)
+
+if __name__ == "__main__":
+    main()

chichi_speech-0.1.0.dist-info/METADATA

@@ -0,0 +1,104 @@
+Metadata-Version: 2.4
+Name: chichi-speech
+Version: 0.1.0
+Summary: Qwen3 TTS Service with Voice Cloning
+Requires-Python: >=3.10
+Requires-Dist: fastapi
+Requires-Dist: numba>=0.59.0
+Requires-Dist: pydantic
+Requires-Dist: qwen-tts
+Requires-Dist: requests
+Requires-Dist: soundfile
+Requires-Dist: torch
+Requires-Dist: uvicorn
+Provides-Extra: dev
+Requires-Dist: build; extra == 'dev'
+Requires-Dist: httpx; extra == 'dev'
+Requires-Dist: pytest; extra == 'dev'
+Description-Content-Type: text/markdown
+
+# Chichi Speech
+
+**A high-quality, voice-cloning TTS service powered by Qwen3.**
+
+Chichi Speech provides a robust REST API and CLI tools for text-to-speech synthesis, featuring efficient voice cloning capabilities. It is designed to be easily deployed or integrated into other AI agents and workflows.
+
+## Features
+
+-   **High Quality**: Utilizes the Qwen3-TTS model for state-of-the-art speech synthesis.
+-   **Voice Cloning**: Clone voices from reference audio files.
+-   **Efficient**: Optimized for reusing voice prompts to minimize computation for repeated requests.
+-   **Standardized API**: Simple REST API (`/synthesize`) for easy integration.
+-   **CLI Tools**: Includes `chichi-speech-server` and `chichi-speech-client` for immediate use.
+
+## Installation
+
+Prerequisites: `git`, `uv`, `python >= 3.10`.
+
+```bash
+export CHICHI_SPEECH_HOME="~/chichi-speech/"
+export CHICHI_SPEECH_ENV="~/chichi-speech/.venv"
+git clone https://github.com/yourusername/chichi-speech.git $CHICHI_SPEECH_HOME
+cd $CHICHI_SPEECH_HOME
+
+uv venv $CHICHI_SPEECH_ENV --python 3.10
+source $CHICHI_SPEECH_ENV/bin/activate
+
+uv pip install -e .
+```
+
+## Usage
+
+### 1. Start the Service
+
+The service runs on port **9090** by default.
+
+```bash
+# Start the server (runs in foreground, use & for background or a separate terminal)
+source $$CHICHI_SPEECH_ENV/bin/activate
+chichi-speech-server
+# OR specify the port explicitly
+chichi-speech-server --port 9090 --host 0.0.0.0
+# OR specify your reference audio and text for voice cloning (Recommended)
+chichi-speech-server --ref-audio /path/to/my/voice.wav --ref-text "caption of the reference audio"
+```
+
+### 2. Verify Service is Running
+Check the health/docs:
+```bash
+curl http://localhost:9090/docs
+```
+
+### 3. Generate Speech
+
+Use cURL:
+```bash
+curl -X POST "http://localhost:9090/synthesize" \
+     -H "Content-Type: application/json" \
+     -d '{
+           "text": "Nice to meet you",
+           "language": "English"
+         }' \
+     --output output/nice_to_meet.wav
+```
+
+## Functionality
+-   **Endpoint**: `POST /synthesize`
+-   **Default Port**: 9090
+-   **Voice Cloning**: Uses a pre-computed voice prompt from reference files to ensure the cloned voice is consistent and generation is fast.
+
+# Development
+
+Install dev dependencies:
+```bash
+uv pip install -e ".[dev]"
+```
+
+Run tests:
+```bash
+pytest
+```
+
+## License
+
+MIT

chichi_speech-0.1.0.dist-info/RECORD

@@ -0,0 +1,7 @@
+chichi_speech/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+chichi_speech/client.py,sha256=GYIxqKy1ONsOxNMx83gMOT1cM6K0rdjWSR_lnPiLk8U,1381
+chichi_speech/server.py,sha256=JnC7A7casb3q_HRot9sSjgA_41E-H7lnJHr_tHOtSMg,4204
+chichi_speech-0.1.0.dist-info/METADATA,sha256=tL48c1POhy-6QQXlmEMbTx4Z3k1GHix-3ivbEZN3-ps,2844
+chichi_speech-0.1.0.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
+chichi_speech-0.1.0.dist-info/entry_points.txt,sha256=tRZ0iWPkkhzyEEQUSnGhyjbFJwxX49JkyqfU5xPR9yM,116
+chichi_speech-0.1.0.dist-info/RECORD,,

chichi_speech-0.1.0.dist-info/WHEEL

@@ -0,0 +1,4 @@
+Wheel-Version: 1.0
+Generator: hatchling 1.28.0
+Root-Is-Purelib: true
+Tag: py3-none-any

chichi_speech-0.1.0.dist-info/entry_points.txt

@@ -0,0 +1,3 @@
+[console_scripts]
+chichi-speech-client = chichi_speech.client:main
+chichi-speech-server = chichi_speech.server:main