Update app.py

app.py

@@ -61,7 +61,6 @@ def convert_file_to_json(file_path: str, file_type: str) -> str:
         if os.path.exists(cache_path):
             with open(cache_path, "r", encoding="utf-8") as f:
                 return f.read()
-
         if file_type == "pdf":
             text = extract_all_pages(file_path)
             result = json.dumps({"filename": os.path.basename(file_path), "content": text, "status": "initial"})
@@ -150,6 +149,7 @@ def init_agent():
         force_finish=True,
         enable_checker=False,
         enable_rag=False,
+        enable_finish=False,  # MODIFIED: Disable Finish tool
         tool_files_dict=None,
         step_rag_num=0,
         seed=100,
@@ -185,7 +185,7 @@ def create_ui(agent):
             chunk_size = 800
             chunks = [extracted[i:i + chunk_size] for i in range(0, len(extracted), chunk_size)]
             chunk_responses = []
-            batch_size = 4
+            batch_size = 8  # MODIFIED: Increase for parallelism
             total_chunks = len(chunks)
 
             prompt_template = """
@@ -204,6 +204,7 @@ Records:
                 for i in range(0, len(chunks), batch_size):
                     batch = chunks[i:i + batch_size]
                     batch_responses = []
+                    log_system_usage(f"Batch {i//batch_size + 1}")  # MODIFIED: Log VRAM
                     for j, chunk in enumerate(batch):
                         prompt = prompt_template.format(chunk=chunk)
                         chunk_response = ""
@@ -211,7 +212,7 @@ Records:
                             message=prompt,
                             history=[],
                             temperature=0.1,
-                            max_new_tokens=96,
+                            max_new_tokens=64,  # MODIFIED: Reduce for speed
                             max_token=4096,
                             call_agent=False,
                             conversation=[],