Update app.py

app.py

@@ -2,20 +2,21 @@ import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 from diffusers import StableDiffusionPipeline
 import torch
-import requests
-from PIL import Image
-import io
-from huggingface_hub import login  # Correct import for authentication
 import os
+import logging
+
+# Set up logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
 
 # Read the Hugging Face access token from the environment variable
 read_token = os.getenv('AccToken')
 if not read_token:
     raise ValueError("Hugging Face access token not found. Please set the AccToken environment variable.")
+from huggingface_hub import login
 login(read_token)
 
 # Define a dictionary of conversational models
-
 conversational_models = {
     "Qwen": "Qwen/QwQ-32B",
     "DeepSeek R1": "deepseek-ai/DeepSeek-R1",
@@ -59,16 +60,15 @@ summarization_pipeline = pipeline("summarization", model="facebook/bart-large-cn
 try:
     text_to_audio_pipeline = pipeline("text-to-audio", model="stabilityai/stable-audio-open-1.0", use_auth_token=read_token)
 except ValueError as e:
-    print(f"Error loading stabilityai/stable-audio-open-1.0: {e}")
-    print("Falling back to a different text-to-audio model.")
+    logger.error(f"Error loading stabilityai/stable-audio-open-1.0: {e}")
+    logger.info("Falling back to a different text-to-audio model.")
     text_to_audio_pipeline = pipeline("text-to-audio", model="microsoft/speecht5_tts")
 
 audio_classification_pipeline = pipeline("audio-classification", model="facebook/wav2vec2-base")
 
-
 def load_conversational_model(model_name):
     if model_name not in conversational_models_loaded:
-        print(f"Loading conversational model: {model_name}")
+        logger.info(f"Loading conversational model: {model_name}")
         tokenizer = AutoTokenizer.from_pretrained(conversational_models[model_name], use_auth_token=read_token)
         model = AutoModelForCausalLM.from_pretrained(conversational_models[model_name], use_auth_token=read_token)
         conversational_tokenizers[model_name] = tokenizer
@@ -97,7 +97,7 @@ def chat(model_name, user_input, history=[]):
 
 def generate_image(model_name, prompt):
     if model_name not in text_to_image_pipelines:
-        print(f"Loading text-to-image model: {model_name}")
+        logger.info(f"Loading text-to-image model: {model_name}")
         text_to_image_pipelines[model_name] = StableDiffusionPipeline.from_pretrained(
             text_to_image_models[model_name], use_auth_token=read_token
         )
@@ -107,7 +107,7 @@ def generate_image(model_name, prompt):
 
 def generate_speech(model_name, text):
     if model_name not in text_to_speech_pipelines:
-        print(f"Loading text-to-speech model: {model_name}")
+        logger.info(f"Loading text-to-speech model: {model_name}")
         text_to_speech_pipelines[model_name] = pipeline(
             "text-to-speech", model=text_to_speech_models[model_name], use_auth_token=read_token
         )