Upload 3 files

Dockerfile

@@ -0,0 +1,21 @@
+# Use lightweight Python image
+FROM python:3.10-slim
+
+# Install system deps
+RUN apt-get update && apt-get install -y git ffmpeg libsndfile1 && rm -rf /var/lib/apt/lists/*
+
+# Set workdir
+WORKDIR /code
+
+# Install requirements
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy app
+COPY app.py .
+
+# Expose port for HF Space
+EXPOSE 7860
+
+# Run with uvicorn
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

app.py

@@ -0,0 +1,29 @@
+from fastapi import FastAPI, Query
+from fastapi.responses import FileResponse
+from transformers import VitsModel, AutoTokenizer
+import torch, scipy.io.wavfile as wavfile
+import uuid, os
+
+app = FastAPI(title="Bambara TTS API")
+
+# Load model once at startup
+model = VitsModel.from_pretrained("facebook/mms-tts-bam")
+tokenizer = AutoTokenizer.from_pretrained("facebook/mms-tts-bam")
+sampling_rate = model.config.sampling_rate
+
+@app.get("/tts/")
+async def tts(text: str = Query(..., description="Bambara text to synthesize")):
+    # Tokenize input
+    inputs = tokenizer(text, return_tensors="pt")
+
+    # Generate waveform
+    with torch.no_grad():
+        output = model(**inputs).waveform
+
+    waveform = output[0]
+
+    # Save to temp file
+    filename = f"tts_{uuid.uuid4().hex}.wav"
+    wavfile.write(filename, rate=sampling_rate, data=waveform.numpy())
+
+    return FileResponse(filename, media_type="audio/wav", filename=filename)

requirements.txt

@@ -0,0 +1,6 @@
+fastapi
+uvicorn
+transformers==4.44.2
+accelerate
+torch
+scipy