app.py

import gradio as gr
from transformers import GPT2Tokenizer, GPT2LMHeadModel, TextDataset, DataCollatorForLanguageModeling
from transformers import Trainer, TrainingArguments
from os.path import dirname
import torch


model_output_path="change"
my_model = GPT2LMHeadModel.from_pretrained(model_output_path)
my_tokenizer = GPT2Tokenizer.from_pretrained(model_output_path)

def generate_response(model, tokenizer, prompt, max_length=200):

    input_ids = tokenizer.encode(prompt, return_tensors="pt")                 
    attention_mask = torch.ones_like(input_ids)
    pad_token_id = tokenizer.eos_token_id

    output = model.generate(
        input_ids,
        max_length=max_length,
        num_return_sequences=1,
        attention_mask=attention_mask,
        pad_token_id=pad_token_id
    )

    return tokenizer.decode(output[0], skip_special_tokens=True)

def predict(prompt):
  response = generate_response(my_model, my_tokenizer, prompt)
  return response

iface = gr.Interface(fn=predict, inputs="text", outputs="text")
iface.launch()