app.py

app.py

@@ -3,7 +3,6 @@ from pathlib import Path
 import pandas as pd
 import numpy as np
 import torch
-import clip
 from PIL import Image
 from io import BytesIO
 import requests
@@ -53,7 +52,7 @@ def encode_search_query(search_query, model, device):
         #text_encoded = model.encode_text(clip.tokenize(search_query).to(device))
         #text_encoded /= text_encoded.norm(dim=-1, keepdim=True)
         # Retrieve the feature vector from the GPU and convert it to a numpy array
-        return model.get_text_features(**inputs).cpu().numpy()
+    return model.get_text_features(**inputs).cpu().numpy()
         #return text_encoded.cpu().numpy()
         
 # Find all matched photos
@@ -87,16 +86,16 @@ def image_search(search_text, search_image, option):
   
     return show_output_image(matched_images)
   elif option == "Image-To-Image":
-    # Input Image for Search
-    with torch.no_grad():
-        processed_image = processor(text=None, images=search_image, return_tensors="pt", padding=True)["pixel_values"]
-        image_feature = model.get_image_features(processed_image.to(device))
-        image_feature /= image_feature.norm(dim=-1, keepdim=True)
-        image_feature = image_feature.cpu().numpy()
+      # Input Image for Search
+      with torch.no_grad():
+         processed_image = processor(text=None, images=search_image, return_tensors="pt", padding=True)["pixel_values"]
+         image_feature = model.get_image_features(processed_image.to(device))
+         image_feature /= image_feature.norm(dim=-1, keepdim=True)
+         image_feature = image_feature.cpu().numpy()
       # Find the matched Images
-        matched_images = find_matches(image_feature, photo_features, photo_ids, 4)
+      matched_images = find_matches(image_feature, photo_features, photo_ids, 4)
       #is_input_image = True
-        return show_output_image(matched_images)
+      return show_output_image(matched_images)
   
 gr.Interface(fn=image_search, 
             inputs=[gr.inputs.Textbox(lines=7, label="Input Text"),