Update app.py

app.py

@@ -7,7 +7,7 @@ from PIL import Image
 from io import BytesIO
 import torch
 import clip
-from gradio_client import Client
+
 
 # Load the segmentation model
 sam_checkpoint = "sam_vit_h_4b8939.pth"
@@ -92,34 +92,13 @@ def segment_image(input_image, text_input):
     # Limit the output to the top 6 key-value pairs
     segmented_regions = segmented_regions[:6]
 
-    # # Run each segmented image through the inference API and store results
-    # results_with_similarity = []
-    # for result in segmented_regions:
-    #     image_base64 = result["image"]
-    #     similarity = result["similarity"]
-        
-    #     # Make an API call with image_base64, get the API result
-    #     client = Client("https://ktllc-clip-model-inputbase64.hf.space/--replicas/dv889/")
-    #     api_result = client.predict(image_base64, text_input, api_name="/predict")
-
-    #     # Append the API result and similarity to the list
-    #     results_with_similarity.append({"api_result": api_result, "image": image_base64, "similarity": similarity})
-
-    #     results_with_similarity.sort(key=lambda x: x["similarity"], reverse=True)
-
-    # results_with_similarity = results_with_similarity[:1]
-        
-    # Print the top 6 results
-    # for result in results_with_similarity:
-    #     print(result)
-    
     # Return the segmented images in descending order of similarity
     return segmented_regions
 
 # Create Gradio components
 input_image = gr.Textbox(label="Base64 Image", lines=8)
 text_input = gr.Textbox(label="Text Input")  # Use Textbox with a label
-output_images = gr.outputs.JSON()
+#output_images = gr.outputs.JSON()
 
 # Create a Gradio interface
-gr.Interface(fn=segment_image, inputs=[input_image, text_input], outputs=output_images).launch()
+gr.Interface(fn=segment_image, inputs=[input_image, text_input], outputs="text").launch()