added input hash; excluding init input from db write

app.py

@@ -106,14 +106,21 @@ def show_demo(placeholder):
                 if 'input_data' in st.session_state:
 
                     if 'yaml_dict' in locals():
+
+                        input_data_serialized = utils.serialize_data(yaml_dict)
+                        input_data_hashed = utils.hash(input_data_serialized)
                         payload = {
                             'user_id': st.session_state.user_id,
                             'timestap': datetime.now().strftime('%Y-%m-%d %H:%M:%S'),
-                            'input_data': utils.serialize_data(yaml_dict)
+                            'input_hash': input_data_hashed,
+                            'input_data': input_data_serialized
                         }
-                        if os.environ.get('remote_model_path'):
-                            # not logging locally
-                            db.write_to_db(st.session_state.db, payload)
+                                                
+                        if st.session_state['init_input_hash'] != input_data_hashed:
+                            # not logging default example
+                            if os.environ.get('remote_model_path'):
+                                # not logging locally
+                                db.write_to_db(st.session_state.db, payload)
 
                     tab1, tab2, tab3 = st.tabs(["Item Correlations", "Scale Correlations", "Scale Reliabilities"])
 
@@ -233,6 +240,10 @@ def main():
         with open('sample_input.yaml', 'r') as file:
             try:
                 st.session_state['input_yaml'] = file.read()
+                init_input_dict = yaml.safe_load(st.session_state['input_yaml'])
+                init_input_serialized = utils.serialize_data(init_input_dict)
+                init_input_hashed = utils.hash(init_input_serialized)
+                st.session_state['init_input_hash'] = init_input_hashed
             except Exception as error:
                 print(error)
 

requirements.txt

@@ -1,4 +1,5 @@
 torch
+hashlib
 pandas==2.2.0
 numpy==1.26.3
 sentence_transformers==2.7.0

utils.py

@@ -1,6 +1,7 @@
 import cleantext
 import ftfy
 import json
+import hashlib
 from io import StringIO
 from datetime import datetime
 
@@ -31,4 +32,10 @@ def serialize_data(data):
         if isinstance(o, datetime):
             return o.__str__()
 
-    return json.dumps(data, default=converter)
+    return json.dumps(data, default=converter)
+
+def hash(input):
+    sha1 = hashlib.sha1()
+    encoded = json.dumps([input], sort_keys=True).encode()
+    sha1.update(encoded)
+    return sha1.hexdigest()