Update app.py

app.py

@@ -15,28 +15,24 @@ from sklearn.preprocessing import normalize, OneHotEncoder
 # loading the train dataset
 with open('clip_train.pkl', 'rb') as f: 
     temp_d = pickle.load(f)
-    # train_xv = temp_d['image'].astype(np.float64)   # Array of image features : np ndarray
-    # train_xt = temp_d['text'].astype(np.float64)    # Array of text features : np ndarray
-    # train_yv  = temp_d['label']                     # Array of labels 
+    train_xv = temp_d['image'].astype(np.float64)   # Array of image features : np ndarray
+    train_xt = temp_d['text'].astype(np.float64)    # Array of text features : np ndarray
+    train_yv  = temp_d['label']                     # Array of labels 
     train_yt  = temp_d['label']                     # Array of labels 
-    # ids = list(temp_d['ids'])                       # image names == len(images)
-
-# train_yt = np.load("train_yt.npy")
+    ids = list(temp_d['ids'])                       # image names == len(images)
 
 # loading the test dataset
 with open('clip_test.pkl', 'rb') as f:
     temp_d = pickle.load(f)
-    # test_xv = temp_d['image'].astype(np.float64)
-    # test_xt = temp_d['text'].astype(np.float64)
-    # test_yv = temp_d['label']
+    test_xv = temp_d['image'].astype(np.float64)
+    test_xt = temp_d['text'].astype(np.float64)
+    test_yv = temp_d['label']
     test_yt = temp_d['label']
 
-# test_xt = np.load("test_xt.npy")
-
 enc = OneHotEncoder(sparse=False)
 enc.fit(np.concatenate((train_yt, test_yt)).reshape((-1, 1)))
-# train_yv = enc.transform(self.train_yv.reshape((-1, 1))).astype(np.float64)
-# test_yv = enc.transform(self.test_yv.reshape((-1, 1))).astype(np.float64)
+train_yv = enc.transform(train_yv.reshape((-1, 1))).astype(np.float64)
+test_yv = enc.transform(test_yv.reshape((-1, 1))).astype(np.float64)
 train_yt = enc.transform(train_yt.reshape((-1, 1))).astype(np.float64)
 test_yt = enc.transform(test_yt.reshape((-1, 1))).astype(np.float64)
 
@@ -55,18 +51,15 @@ d = 32
 text_index = faiss.index_factory(d, "Flat", faiss.METRIC_INNER_PRODUCT)
 text_index = faiss.read_index("text_index.index")
 
-def T2Isearch(query,focussed_word, k=50):
+def T2Isearch(query, k=50):
     # Encode the text query        
-    inputs = text_tokenizer([query,focussed_word], padding=True, return_tensors="pt")    
+    inputs = text_tokenizer([query], padding=True, return_tensors="pt")    
     outputs = text_model(**inputs)
     query_embedding = outputs.text_embeds
     query_vector = query_embedding.detach().numpy()
-    query_vector = np.concatenate((query_vector[0], query_vector[1]), dtype=np.float32)
-    query_vector = query_vector.reshape(1,1024)
+    query_vector = query_vector.reshape(1,512)
     faiss.normalize_L2(query_vector)
-    text_index.nprobe = text_index.ntotal
-
-    # text_index.nprobe = 100
+    index.nprobe = index.ntotal
     
     # Search for the nearest neighbors in the FAISS text index
     D, I = text_index.search(query_vector, k)
@@ -104,7 +97,7 @@ def T2Isearch(query,focussed_word, k=50):
             if count == 5: break
 
 query = st.text_input("Enter your search query here:")
-focussed_word = st.text_input("Enter focussed word here")
+
 if st.button("Search"):
     if query:
-        T2Isearch(query, focussed_word)
+        T2Isearch(query)