app.py

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
import gradio as gr


model = AutoModelForCausalLM.from_pretrained("MohamedTalaat91/gpt2-wikitext2")
tokenizer = AutoTokenizer.from_pretrained("MohamedTalaat91/gpt2-tokenizer")



def generate(input_text) :
    inputs = tokenizer(input_text, return_tensors="pt")
    # Generate text based on the input
    generated_ids = model.generate(
        inputs['input_ids'], 
        max_length=100,  # Adjust the max length as needed
        num_return_sequences=1,  # Number of texts to generate
        do_sample=True,  # Enable sampling (as opposed to greedy search)
        top_k=50,  # Top-k sampling to introduce diversity
        temperature=0.7  # Controls randomness in sampling
    )
    generated_text = tokenizer.decode(generated_ids[0], skip_special_tokens=True)

    return generated_text


import gradio as gr

gr.ChatInterface(generate, type="messages").launch()