Add Respite API discovery metadata

app.py

@@ -1,9 +1,13 @@
 import os
+import platform
+import shutil
 import tempfile
 import torch
 import torchaudio
 import gradio as gr
 import spaces
+from fastapi import FastAPI
+from fastapi.responses import JSONResponse
 from einops import rearrange
 from huggingface_hub import login
 from stable_audio_3 import StableAudioModel
@@ -27,6 +31,89 @@ except Exception as e:
     print(f"Warning: ZeroGPU touch failed (running CPU-only): {e}")
 
 
+# API metadata
+API_RESOURCES = {
+    "audio_generation": {
+        "name": "Audio generation",
+        "description": "Generate music or sound effects from a text prompt.",
+        "endpoint": "/api/audio/generate",
+        "method": "POST",
+        "input": {
+            "prompt": "string",
+            "duration": "number (1-120 seconds)",
+            "steps": "integer (1-50)",
+            "cfg_scale": "number (0-10)",
+            "seed": "integer (-1 for random)",
+            "model": "small-music | small-sfx"
+        },
+        "output": "WAV audio file"
+    }
+}
+
+def get_server_specs():
+    storage = shutil.disk_usage(os.getcwd())
+    return {
+        "name": "Respite API",
+        "version": "1.0.0",
+        "description": "General-purpose AI API server with audio generation capabilities.",
+        "base_path": "/api",
+        "authentication": "none",
+        "content_types": ["application/json", "audio/wav"],
+        "resources_endpoint": "/api/resources",
+        "specs_endpoint": "/api/specs",
+        "runtime": {
+            "platform": platform.platform(),
+            "python_version": platform.python_version(),
+            "cpu_cores": os.cpu_count(),
+            "ram_bytes": _get_ram_bytes(),
+            "storage_total_bytes": storage.total,
+            "storage_used_bytes": storage.used,
+            "storage_free_bytes": storage.free
+        },
+        "limits": {
+            "max_concurrent_requests": 1,
+            "max_queue_size": 4,
+            "audio_max_duration_seconds": 120
+        }
+    }
+
+
+def _get_ram_bytes():
+    try:
+        with open("/proc/meminfo", "r", encoding="utf-8") as meminfo:
+            for line in meminfo:
+                if line.startswith("MemTotal:"):
+                    return int(line.split()[1]) * 1024
+    except (FileNotFoundError, OSError, ValueError):
+        pass
+    return None
+
+
+API_SPECS = {
+    "name": "Respite API",
+    "version": "1.0.0",
+    "description": "General-purpose AI API server with audio generation capabilities.",
+    "base_path": "/api",
+    "authentication": "none",
+    "content_types": ["application/json", "audio/wav"],
+    "resources_endpoint": "/api/resources",
+    "specs_endpoint": "/api/specs",
+    "limits": {
+        "max_concurrent_requests": 1,
+        "max_queue_size": 4,
+        "audio_max_duration_seconds": 120
+    }
+}
+
+
+def get_resources():
+    return {"resources": API_RESOURCES}
+
+
+def get_specs():
+    return {**API_SPECS, "runtime": get_server_specs()["runtime"]}
+
+
 # Model cache
 MODEL_CACHE = {}
 
@@ -81,6 +168,19 @@ def generate_audio(prompt, duration, steps, cfg_scale, seed, model_name):
     return output_path
 
 
+api = FastAPI(title=API_SPECS["name"], version=API_SPECS["version"])
+
+
+@api.get("/api/resources")
+def resources():
+    return JSONResponse(get_resources())
+
+
+@api.get("/api/specs")
+def specs():
+    return JSONResponse(get_specs())
+
+
 with gr.Blocks(title="Stable Audio 3 Small") as demo:
     gr.Markdown("# 🎵 Stable Audio 3 Small - Music & SFX Generation")
     gr.Markdown("Generate music and sound effects using Stability AI's Stable Audio 3 Small models. Runs on CPU.")
@@ -127,4 +227,6 @@ with gr.Blocks(title="Stable Audio 3 Small") as demo:
     )
 
 
-demo.queue(max_size=4, default_concurrency_limit=1).launch()
+demo.queue(max_size=4, default_concurrency_limit=1).launch(
+    app_kwargs={"app": api}
+)