Update pages/type_text.py

pages/type_text.py

@@ -103,26 +103,26 @@ numMAPPINGS_input = 5
 #numMAPPINGS_input = st.text_input("Type number of mappings and hit Enter", key="user_input_numMAPPINGS")
 #st.button("Clear text", on_click=on_click)
 
-SentenceTransformer = st.selectbox(
-    "Select the Sentence Transformer model",
-    ("all-MiniLM-L6-v2 (original model for general domain)", "all-MiniLM-L6-v2 (fine-tuned model for medical domain)",
-     "all-mpnet-base-v2 (original model for general domain)", "all-mpnet-base-v2 (fine-tuned model for medical domain)")
-)
-st.write("You selected:", SentenceTransformer)
-
-
-
-
-
-
-
-
-
-
 
+## Define the SentTrans models
+st_models = {
+    'original model for general domain, fastest: all-MiniLM-L6-v2': 'all-MiniLM-L6-v2', 
+    'fine-tuned model for medical domain: all-MiniLM-L6-v2': 'all-MiniLM-L6-v2',
+    'original model for general domain, best performance: all-mpnet-base-v2': 'all-mpnet-base-v2', 
+    'fine-tuned model for medical domain: all-mpnet-base-v2': 'all-mpnet-base-v2',
+}
+ 
+## Create the select box
+selected_st_model = st.selectbox('Choose a model:', list(st_models.keys()))
+st.write("You selected:", selected_st_model)
+
+## Get the selected model
+SentTrans_model = st_models[selected_st_model]
+ 
+## Use the model...
 @st.cache_resource
 def load_model():
-    model = SentenceTransformer('all-MiniLM-L6-v2') # fastest
+    model = SentenceTransformer(SentTrans_model)
     return model
 model = load_model() 
 
@@ -136,14 +136,6 @@ model = load_model()
 
 
 
-
-
-
-
-
-
-
-
 INTdesc_embedding = model.encode(INTdesc_input)
 
 # Semantic search, Compute cosine similarity between INTdesc_embedding and SBS descriptions