update

app.py

@@ -17,6 +17,8 @@ if hf_token:
 else:
     logger.error("HF_TOKEN not found in environment variables")
 
+SAMPLE_NUM = 10
+
 # Set up logging
 logging.basicConfig(level=logging.INFO,
                     format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
@@ -79,15 +81,17 @@ def save_session_data(username, data):
 
         logger.info(f"Data uploaded to Hugging Face Data Space:\n{json_data}")
 
-
-        # Upload the file to a separate Hugging Face Space dedicated to data storage
-        api = HfApi()
-        api.upload_file(
-            path_or_fileobj=temp_file_path,
-            path_in_repo=f"session_data_pref/{file_name}",
-            repo_id="luulinh90s/Tabular-LLM-Study-Data",
-            repo_type="space",
-        )
+        if json_data['current_index'] < SAMPLE_NUM:
+            pass
+        else:
+            # Upload the file to a separate Hugging Face Space dedicated to data storage
+            api = HfApi()
+            api.upload_file(
+                path_or_fileobj=temp_file_path,
+                path_in_repo=f"session_data_pref/{file_name}",
+                repo_id="luulinh90s/Tabular-LLM-Study-Data",
+                repo_type="space",
+            )
 
         # Remove the temporary file
         os.remove(temp_file_path)
@@ -154,14 +158,14 @@ def load_samples(methods):
 
 def select_balanced_samples(samples):
     try:
-        # Ensure we have at least 10 unique samples
+        # Ensure we have at least N unique samples
         unique_samples = list({(s['category'], s['file']) for s in samples})
 
-        if len(unique_samples) < 10:
+        if len(unique_samples) < SAMPLE_NUM:
             logger.warning(f"Not enough unique samples. Only {len(unique_samples)} available.")
             selected_samples = unique_samples
         else:
-            selected_samples = random.sample(unique_samples, 10)
+            selected_samples = random.sample(unique_samples, SAMPLE_NUM)
 
         logger.info(f"Unique sampled samples:\n{selected_samples}")