Upload 3 files

Dockerfile

@@ -0,0 +1,10 @@
+FROM pytorch/pytorch:2.1.0-cuda11.8-cudnn8-runtime
+
+WORKDIR /app
+
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY . .
+
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

app.py

@@ -0,0 +1,39 @@
+from fastapi import FastAPI
+import torch
+import uuid
+import os
+import soundfile as sf
+
+from diffusers import StableAudioPipeline
+
+app = FastAPI()
+
+OUTPUT_DIR = "outputs"
+os.makedirs(OUTPUT_DIR, exist_ok=True)
+
+print("Loading Stable Audio Open...")
+pipe = StableAudioPipeline.from_pretrained(
+    "stabilityai/stable-audio-open-1.0",
+    torch_dtype=torch.float16
+)
+pipe.to("cuda")
+
+@app.post("/generate")
+async def generate(prompt: str, duration: float = 10.0):
+    audio = pipe(
+        prompt,
+        duration=duration,
+        guidance_scale=7.5,
+        num_inference_steps=200
+    ).audios[0]
+
+    filename = f"{uuid.uuid4().hex}.wav"
+    path = os.path.join(OUTPUT_DIR, filename)
+
+    sf.write(path, audio, samplerate=44100)
+
+    return {
+        "prompt": prompt,
+        "duration": duration,
+        "file": filename
+    }

requirements.txt

@@ -0,0 +1,9 @@
+fastapi
+uvicorn
+torch
+torchaudio
+diffusers
+transformers
+accelerate
+soundfile
+scipy