Update app.py

app.py

@@ -2,13 +2,16 @@ import os
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForSequenceClassification
 import torch
+import warnings
+
+# Suppress NVML warning
+warnings.filterwarnings("ignore", message="Can't initialize NVML")
 
 os.environ["TRANSFORMERS_CACHE"] = "/tmp"
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 if torch.cuda.is_available():
     os.environ["CUDA_VISIBLE_DEVICES"] = "0"
-    torch.cuda.init()
 
 MODEL_NAME = "s-nlp/roberta-base-formality-ranker"
 tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
@@ -17,9 +20,7 @@ model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME)
 model = model.to(device)
 
 def calculate_formality_percentages(score):
-    # Convert score to grayscale percentage (0-100)
     grayscale = int(score * 100)
-    # Use grayscale to determine formal/informal percentages
     formal_percent = grayscale
     informal_percent = 100 - grayscale
     return formal_percent, informal_percent
@@ -37,7 +38,6 @@ def predict_formality(text):
     # Calculate percentages using grayscale
     formal_percent, informal_percent = calculate_formality_percentages(score)
     
-    # Create response in the new format
     response = {
         "formality_score": round(score, 3),
         "formal_percent": formal_percent,
@@ -47,7 +47,6 @@ def predict_formality(text):
     
     return response
 
-# Create Gradio interface
 demo = gr.Interface(
     fn=predict_formality,
     inputs=gr.Textbox(label="Enter your text", lines=3),
@@ -63,9 +62,9 @@ demo = gr.Interface(
 
 # Launch the app
 if __name__ == "__main__":
+    demo.queue() 
     demo.launch(
         server_name="0.0.0.0",
         server_port=7860,
-        share=True, 
-        enable_queue=True  
+        share=True  
     )