app.py

import gradio as gr
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = 'openai-community/gpt2-large'
model = AutoModelForCausalLM.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=True)

def generate_blogpost(topic, max_length=500, temperature=0.7):
    prompt = f"Write a blog post about {topic}:\n\n"

    # Encode input:
    inputs_encoded = tokenizer(prompt, return_tensors='pt')
    # Model Output:
    model_output = model.generate(
        inputs_encoded["input_ids"],
        max_new_tokens=max_length,
        do_sample=True,
        temperature=temperature
    )[0]
    # Decode the output
    output = tokenizer.decode(model_output, skip_special_tokens=True)
    
    # Remove the prompt from the generated text
    blog_post = output[len(prompt):].strip()
    
    return blog_post

# Create the Gradio interface
iface = gr.Interface(
    fn=generate_blogpost,
    inputs=[
        gr.Textbox(lines=1, placeholder="Enter the blog post topic here..."),
        gr.Slider(minimum=100, maximum=1000, step=50, label="Max Length", value=500),
        gr.Slider(minimum=0.1, maximum=1.0, step=0.1, label="Temperature", value=0.7)
    ],
    outputs="text",
    title="GPT2 Blog Post Generator",
    description="Enter a topic, and this app will generate a blog post using GPT-2."
)

# Launch the app
iface.launch()