init

coquitts/Dockerfile

@@ -0,0 +1,18 @@
+FROM python:3.11-slim
+ENV PYTHONUNBUFFERED=1
+WORKDIR /app
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+	ffmpeg \
+	libsndfile1 \
+	build-essential \
+	git \
+	wget \
+ && rm -rf /var/lib/apt/lists/*
+
+COPY requirements.txt ./
+RUN python -m pip install --upgrade pip setuptools wheel && \
+	pip install --no-cache-dir -r requirements.txt
+
+COPY server.py ./
+CMD ["uvicorn", "server:app", "--host", "0.0.0.0", "--port", "8002"]

coquitts/requirements.txt

@@ -0,0 +1,6 @@
+fastapi==0.100.0
+uvicorn[standard]==0.22.0
+TTS==0.12.0
+torch==2.2.0
+soundfile==0.12.1
+numpy==1.26.0

coquitts/server.py

@@ -0,0 +1,28 @@
+from fastapi import FastAPI, HTTPException
+from fastapi import Body
+from fastapi.responses import FileResponse, JSONResponse
+from TTS.api import TTS
+import tempfile
+import os
+
+app = FastAPI()
+
+# Load a German-capable model. Model may be downloaded on first run.
+tts = TTS(model_name="tts_models/de/thorsten_hsmm")
+
+
+@app.post("/speak")
+def speak(payload: dict = Body(...)):
+    text = payload.get("text")
+    language = payload.get("language", "de")
+    if not text:
+        raise HTTPException(status_code=400, detail="Missing 'text' in body")
+
+    fd, path = tempfile.mkstemp(suffix=".wav")
+    os.close(fd)
+    try:
+        tts.tts_to_file(text=text, speaker=None, language=language, file_path=path)
+        return FileResponse(path, media_type="audio/wav", filename="response.wav")
+    finally:
+        # FileResponse will stream the file; don't remove immediately. Consumer can manage cleanup.
+        pass