Update app.py

app.py

@@ -1,6 +1,3 @@
-import warnings
-warnings.filterwarnings("ignore")
-
 import os
 import sys
 from typing import List, Tuple
@@ -12,8 +9,6 @@ from llama_cpp_agent.chat_history.messages import Roles
 from llama_cpp_agent.messages_formatter import MessagesFormatter, PromptMarkers
 from huggingface_hub import hf_hub_download
 import gradio as gr
-from logger import logging
-from exception import CustomExceptionHandling
 
 # Load the Environment Variables from .env file
 huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
@@ -97,69 +92,65 @@ def respond(
     Yields:
         str: The translated text as it is generated.
     """
-    try:
-        global llm, llm_model
-        if llm is None or llm_model != model:
-            model_path = f"models/{model}"
-            if not os.path.exists(model_path):
-                yield f"Error: Model file not found at {model_path}."
-                return
-            llm = Llama(
-                model_path=model_path,
-                flash_attn=False,
-                n_gpu_layers=0,
-                n_batch=8,
-                n_ctx=2048,
-                n_threads=8,
-                n_threads_batch=8,
-            )
-            llm_model = model
-        provider = LlamaCppPythonProvider(llm)
-
-        # Get system prompt and user prefix based on direction
-        prompts = direction_to_prompts[direction]
-        system_message = prompts["system"]
-        user_prefix = prompts["prefix"]
-
-        agent = LlamaCppAgent(
-            provider,
-            system_prompt=system_message,
-            custom_messages_formatter=gemma_3_formatter,
-            debug_output=True,
-        )
-
-        settings = provider.get_provider_default_settings()
-        settings.temperature = temperature
-        settings.top_k = top_k
-        settings.top_p = top_p
-        settings.max_tokens = max_tokens
-        settings.repeat_penalty = repeat_penalty
-        settings.stream = True
-
-        messages = BasicChatHistory()
-        for user_msg, assistant_msg in history:
-            full_user_msg = user_prefix + " " + user_msg
-            messages.add_message({"role": Roles.user, "content": full_user_msg})
-            messages.add_message({"role": Roles.assistant, "content": assistant_msg})
-
-        full_message = user_prefix + " " + message
-
-        stream = agent.get_chat_response(
-            full_message,
-            llm_sampling_settings=settings,
-            chat_history=messages,
-            returns_streaming_generator=True,
-            print_output=False,
+    
+    global llm, llm_model
+    if llm is None or llm_model != model:
+        model_path = f"models/{model}"
+        if not os.path.exists(model_path):
+            yield f"Error: Model file not found at {model_path}."
+            return
+        llm = Llama(
+            model_path=model_path,
+            flash_attn=False,
+            n_gpu_layers=0,
+            n_batch=8,
+            n_ctx=2048,
+            n_threads=8,
+            n_threads_batch=8,
         )
-
-        logging.info("Response stream generated successfully")
-        outputs = ""
-        for output in stream:
-            outputs += output
-            yield outputs
-
-    except Exception as e:
-        raise CustomExceptionHandling(e, sys) from e
+        llm_model = model
+    provider = LlamaCppPythonProvider(llm)
+
+    # Get system prompt and user prefix based on direction
+    prompts = direction_to_prompts[direction]
+    system_message = prompts["system"]
+    user_prefix = prompts["prefix"]
+
+    agent = LlamaCppAgent(
+        provider,
+        system_prompt=system_message,
+        custom_messages_formatter=gemma_3_formatter,
+        debug_output=True,
+    )
+
+    settings = provider.get_provider_default_settings()
+    settings.temperature = temperature
+    settings.top_k = top_k
+    settings.top_p = top_p
+    settings.max_tokens = max_tokens
+    settings.repeat_penalty = repeat_penalty
+    settings.stream = True
+
+    messages = BasicChatHistory()
+    for user_msg, assistant_msg in history:
+        full_user_msg = user_prefix + " " + user_msg
+        messages.add_message({"role": Roles.user, "content": full_user_msg})
+        messages.add_message({"role": Roles.assistant, "content": assistant_msg})
+
+    full_message = user_prefix + " " + message
+
+    stream = agent.get_chat_response(
+        full_message,
+        llm_sampling_settings=settings,
+        chat_history=messages,
+        returns_streaming_generator=True,
+        print_output=False,
+    )
+
+    outputs = ""
+    for output in stream:
+        outputs += output
+        yield outputs
 
 demo = gr.ChatInterface(
     respond,