app.py

import torch
import gradio as gr
import requests
import os 
import transformers
from transformers import AutoModelWithLMHead, AutoTokenizer, AutoModelForCausalLM, pipeline
from transformers import GPT2Tokenizer, GPT2Model

tokenizer = AutoTokenizer.from_pretrained("href/gpt2-schiappa")

model = AutoModelForCausalLM.from_pretrained("href/gpt2-schiappa")
pipe = pipeline('text-generation', model="href/gpt2-schiappa", tokenizer=tokenizer)

def text_generation(input_text, seed, min_length, max_length, temperature, top_k, top_p, repetition_penalty):
  input_ids = tokenizer(input_text, return_tensors="pt").input_ids
  torch.manual_seed(seed) # Max value: 18446744073709551615
  outputs = model.generate(input_ids, do_sample=True, max_length=max_length, min_length=min_length, temperature=temperature, top_k=top_k, top_p=top_p, repetition_penalty=repetition_penalty)
  generated_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
  return generated_text

title = "Minelli"
description = "Écrit comme Marie Minelli"

gr.Interface(
    text_generation,
    [
      gr.inputs.Textbox(lines=2, label="Prompt"),
      gr.inputs.Number(default=10, label="Enter seed number"),
      gr.inputs.Slider(default=0, label="Minimum Length", minimum=0),
      gr.inputs.Slider(default=100, label="Maximum Length", minimum=0),
      gr.inputs.Slider(default=1.0, label="Temperature", minimum=0.1),
      gr.inputs.Slider(default=40, label="topK", minimum=1),
      gr.inputs.Slider(default=0.9, label="topP", minimum=0.0, maximum=1.0),
      gr.inputs.Slider(default=1.0, label="Repetition penalty", minimum=1.0)
    ],
    [gr.outputs.Textbox(type="auto", label="Generated")],
    title=title,
    description=description,
    theme="huggingface"
).launch()