app.py

from transformers import AutoModelForCausalLM, AutoTokenizer, AutoConfig
import gradio as gr
import torch


title = "????AI ChatBot bajo GPU"
description = "A State-of-the-Art Large-scale Pretrained Response generation model (DialoGPT)"
examples = [["How are you?"]]
model_id="clibrain/Llama-2-13b-ft-instruct-es-gptq-4bit"
config = AutoConfig.from_pretrained(model_id)
#config.quantization_config["use_exllama"] = True
config.quantization_config["disable_exllama"] = True
config.quantization_config["exllama_config"] = {"version":2}
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
print("********************")
print(device)
print("********************")

model = AutoModelForCausalLM.from_pretrained(model_id, low_cpu_mem_usage=True, config=config)
model = model.to(device)

tokenizer = AutoTokenizer.from_pretrained(model_id)


def predict(input, history=[]):
    # tokenize the new input sentence
    new_user_input_ids = tokenizer.encode(
        input + tokenizer.eos_token, return_tensors="pt"
    ).to(device)

    # append the new user input tokens to the chat history
    historygpu=torch.LongTensor(history).to(device)
    bot_input_ids = torch.cat([historygpu, new_user_input_ids], dim=-1)

    # generate a response
    history = model.generate(
        bot_input_ids, max_length=4000, pad_token_id=tokenizer.eos_token_id
    )
    
    # convert the tokens to text, and then split the responses into lines
    response = tokenizer.decode(history[0]).split("<|endoftext|>")
    # print('decoded_response-->>'+str(response))
    print(response)
    response = [
        (response[i], response[i + 1]) for i in range(0, len(response) - 1, 2)
    ]  # convert to tuples of list
    # print('response-->>'+str(response))
    return response, history


gr.Interface(
    fn=predict,
    title=title,
    description=description,
    examples=examples,
    inputs=["text", "state"],
    outputs=["chatbot", "state"],
    theme="finlaymacklon/boxy_violet",
).launch()