Update app.py

app.py

@@ -2,7 +2,7 @@ import logging
 import gradio as gr
 from queue import Queue
 import time
-from prometheus_client import start_http_server, Counter, Histogram
+from prometheus_client import start_http_server, Counter, Histogram, Gauge
 import threading
 import psutil
 import random
@@ -14,6 +14,11 @@ ner_pipeline = pipeline("ner", model="Sevixdd/roberta-base-finetuned-ner")
 # --- Prometheus Metrics Setup ---
 REQUEST_COUNT = Counter('gradio_request_count', 'Total number of requests')
 REQUEST_LATENCY = Histogram('gradio_request_latency_seconds', 'Request latency in seconds')
+ERROR_COUNT = Counter('gradio_error_count', 'Total number of errors')
+RESPONSE_SIZE = Histogram('gradio_response_size_bytes', 'Size of responses in bytes')
+CPU_USAGE = Gauge('system_cpu_usage_percent', 'System CPU usage in percent')
+MEM_USAGE = Gauge('system_memory_usage_percent', 'System memory usage in percent')
+QUEUE_LENGTH = Gauge('chat_queue_length', 'Length of the chat queue')
 
 # --- Logging Setup ---
 logging.basicConfig(filename="chat_log.txt", level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
@@ -30,15 +35,29 @@ def chat_function(message, history):
             chat_queue.put(message)
             logging.info(f"User: {message}")
 
-            ner_result = ner_pipeline(message)
-            response = f"Response from NER model: {ner_result}"
+            ner_results = ner_pipeline(message)
+
+            detailed_response = []
+            for result in ner_results:
+                token = result['word']
+                score = result['score']
+                entity = result['entity']
+                start = result['start']
+                end = result['end']
+                detailed_response.append(f"Token: {token}, Entity: {entity}, Score: {score:.4f}, Start: {start}, End: {end}")
+
+            response = "\n".join(detailed_response)
             logging.info(f"Bot: {response}")
 
+            response_size = len(response.encode('utf-8'))
+            RESPONSE_SIZE.observe(response_size)
+
             time.sleep(random.uniform(0.5, 2.5))  # Simulate processing time
 
             chat_queue.get()
             return response
         except Exception as e:
+            ERROR_COUNT.inc()
             logging.error(f"Error in chat processing: {e}")
             return "An error occurred. Please try again."
 
@@ -84,6 +103,8 @@ body {
         while True:
             cpu_usage_display.value = psutil.cpu_percent()
             mem_usage_display.value = psutil.virtual_memory().percent
+            CPU_USAGE.set(psutil.cpu_percent())
+            MEM_USAGE.set(psutil.virtual_memory().percent)
             time.sleep(5)
 
     def update_logs(logs_display):
@@ -100,12 +121,18 @@ body {
             model_params_display.value = model_params_str
             time.sleep(10)  # Update every 10 seconds
 
+    def update_queue_length():
+        while True:
+            QUEUE_LENGTH.set(chat_queue.qsize())
+            time.sleep(1)  # Update every second
+
     # --- Start Threads ---
     threading.Thread(target=start_http_server, args=(8000,), daemon=True).start()
     threading.Thread(target=update_metrics, args=(request_count_display, avg_latency_display), daemon=True).start()
     threading.Thread(target=update_usage, args=(cpu_usage_display, mem_usage_display), daemon=True).start()
     threading.Thread(target=update_logs, args=(logs_display,), daemon=True).start()
     threading.Thread(target=display_model_params, args=(model_params_display,), daemon=True).start()
+    threading.Thread(target=update_queue_length, daemon=True).start()
 
     # --- Simulate Chat Interactions ---
     def simulate_interactions():
@@ -118,4 +145,3 @@ body {
 
 # Launch the app
 demo.launch(share=True)
-