Update app.py

app.py

@@ -1,19 +1,19 @@
 import gradio as gr
-from transformers import pipeline, BartTokenizer, BartForConditionalGeneration
+from transformers import pipeline, GPT2Tokenizer, GPT2LMHeadModel
 
 # Initialize the image classification pipeline
 classifier = pipeline("image-classification", model="google/vit-base-patch16-224")
 
 # Initialize the tokenizer and model for the generative text (GPT-like model)
-model_name = "textattack/facebook-bart-base-RTE"  # Example BART model for demonstration
-tokenizer = BartTokenizer.from_pretrained(model_name)
-model = BartForConditionalGeneration.from_pretrained(model_name)
+model_name = "gpt2"  # Use GPT-2 model for demonstration
+tokenizer = GPT2Tokenizer.from_pretrained(model_name)
+model = GPT2LMHeadModel.from_pretrained(model_name)
 
 def generate_tweet(label):
     # Craft a prompt that naturally encourages engaging and relevant tweet content
     prompt = f"write a tweet about {label}"
 
-    inputs = tokenizer.encode(prompt, return_tensors="pt")
+    inputs = tokenizer.encode(prompt, return_tensors="pt", add_special_tokens=True)
     outputs = model.generate(inputs, max_length=280, num_return_sequences=1, no_repeat_ngram_size=2)
 
     tweet = tokenizer.decode(outputs[0], skip_special_tokens=True)
@@ -31,7 +31,7 @@ def predict(image):
     return tweet
 
 title = "Image Classifier to Generative Tweet"
-description = "This demo recognizes and classifies images using the 'google/vit-base-patch16-224' model and generates a tweet about the top prediction using the 'facebook/bart-large-cnn' generative text model."
+description = "This demo recognizes and classifies images using the 'google/vit-base-patch16-224' model and generates a tweet about the top prediction using GPT-2 for generating creative and engaging content."
 input_component = gr.Image(type="pil", label="Upload an image here")
 output_component = gr.Textbox(label="Generated Promotional Tweet")