Create app.py

app.py

@@ -0,0 +1,50 @@
+import gradio as gr
+from transformers import pipeline
+
+# Initialize the text generation pipeline with caching
+# Using a smaller model for better performance on free tier
+generator = pipeline('text-generation', model='gpt2')
+
+def generate_marketing_text(prompt, content_type, max_length=150, temperature=0.7):
+    """
+    Generate marketing text based on the prompt and content type using GPT-2
+    """
+    # Set a reasonable max_length to avoid timeouts
+    if max_length > 250:
+        max_length = 250
+    
+    # Enhance the prompt based on content type
+    enhanced_prompt = f"Write a {content_type} about {prompt}. "
+    
+    # Generate text
+    result = generator(
+        enhanced_prompt, 
+        max_length=max_length, 
+        temperature=temperature, 
+        do_sample=True,
+        pad_token_id=50256
+    )
+    
+    # Return the generated text
+    return result[0]['generated_text']
+
+# Create Gradio interface
+demo = gr.Interface(
+    fn=generate_marketing_text,
+    inputs=[
+        gr.Textbox(lines=3, placeholder="Enter your topic here...", label="Topic"),
+        gr.Radio(
+            ["social media post", "email newsletter", "product description", "ad copy"],
+            label="Content Type",
+            value="social media post"
+        ),
+        gr.Slider(minimum=50, maximum=250, value=150, step=10, label="Maximum Length"),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.7, step=0.1, label="Creativity (Temperature)")
+    ],
+    outputs=gr.Textbox(lines=10, label="Generated Marketing Content"),
+    title="AdGenAI - Marketing Content Generator",
+    description="Enter a topic and select content type to generate marketing content using AI."
+)
+
+# Launch the app with caching enabled for better performance
+demo.launch(cache_examples=True)