Update app.py

app.py

@@ -1,27 +1,39 @@
+from flask import Flask, request, jsonify, send_file
+from flask_cors import CORS  # Import CORS
+from gradio_client import Client
 import gradio as gr
-from random import randint
-from all_models import models
-from externalmod import gr_Interface_load
-import asyncio
 import os
+import asyncio
 from threading import RLock
+from all_models import models  # Import your models
+from externalmod import gr_Interface_load  # Import the model loading function
 
-lock = RLock()
+app = Flask(__name__)
+CORS(app)  # Enable CORS for all routes
+
+# Environment variable for Hugging Face token
 HF_TOKEN = os.environ.get("HF_TOKEN")
+lock = RLock()
+
+# Initialize the Gradio client
+client = Client("Geek7/mdztxi2")
 
+# Load models using gr_Interface_load
 def load_fn(models):
     global models_load
     models_load = {}
-    
+
     for model in models:
         if model not in models_load.keys():
             try:
+                # Use your custom model loading function
                 m = gr_Interface_load(f'models/{model}', hf_token=HF_TOKEN)
             except Exception as error:
                 print(error)
-                m = gr.Interface(lambda: None, ['text'], ['image'])
+                m = gr.Interface(lambda: None, ['text'], ['image'])  # Fallback to a dummy interface
             models_load.update({model: m})
 
+# Load the models
 load_fn(models)
 
 num_models = 6
@@ -30,31 +42,33 @@ default_models = models[:num_models]
 inference_timeout = 600
 
 async def infer(model_str, prompt, seed=1, timeout=inference_timeout):
-    kwargs = {"seed": seed}
-    task = asyncio.create_task(asyncio.to_thread(models_load[model_str].fn, prompt=prompt, **kwargs, token=HF_TOKEN))
-    await asyncio.sleep(0)
     try:
-        result = await asyncio.wait_for(task, timeout=timeout)
-    except (Exception, asyncio.TimeoutError) as e:
-        print(e)
-        print(f"Task timed out: {model_str}")
-        if not task.done(): 
-            task.cancel()
-        result = None
-    if task.done() and result is not None:
+        # Use the gradio_client to make the API call
+        result = await asyncio.to_thread(client.predict, model_str=model_str, prompt=prompt, seed=seed, api_name="/predict")
         with lock:
             png_path = "image.png"
-            result.save(png_path)
+            result.save(png_path)  # Save the image to a file
         return png_path
-    return None
-
-# Expose Gradio API
-def generate_api(model_str, prompt, seed=1):
-    result = asyncio.run(infer(model_str, prompt, seed,api_name="/predict" ))
-    if result:
-        return result  # Path to generated image
-    return None
-
-# Launch Gradio API without frontend
-iface = gr.Interface(fn=generate_api, inputs=["text", "text", "number"], outputs="file")
-iface.launch(show_api=True, share=True)
+    except Exception as e:
+        print(f"Error during inference: {e}")
+        return None
+
+@app.route('/generate-image', methods=['POST'])
+def generate_image():
+    data = request.get_json()
+    model_str = data['model_str']
+    prompt = data['prompt']
+    seed = data['seed']
+
+    # Make a prediction request
+    try:
+        result_path = asyncio.run(infer(model_str, prompt, seed))
+        if result_path:
+            return send_file(result_path, mimetype='image/png')  # Send the generated image file
+        else:
+            return jsonify({"error": "Image generation failed."}), 500
+    except Exception as e:
+        return jsonify({"error": str(e)}), 500
+
+if __name__ == '__main__':
+    app.run(debug=True)