Update app.py

app.py

@@ -1,12 +1,12 @@
 import os
 import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
-import torch
 import spaces
 
 # Create the necessary directories
 os.makedirs('.gradio/cached_examples/17', exist_ok=True)
 
+
 def get_model_name(language):
     """Map language choice to the corresponding model."""
     model_mapping = {
@@ -53,6 +53,17 @@ def generate_kids_story(character, setting, language):
     output = generator(messages)
     return output[0]["generated_text"]
 
+import gc 
+import torch
+# Delete model and associated objects
+del model 
+del tokenizer 
+del generator
+# Run garbage collection
+gc.collect ()
+# Empty CUDA cache
+torch.cuda.empty_cache()
+
 # Create Gradio interface
 demo = gr.Interface(
     fn=generate_kids_story,
@@ -78,4 +89,4 @@ demo = gr.Interface(
 )
 
 # Launch the Gradio app
-demo.launch(cache_examples = False)
+demo.launch()