app.py

# demo5
# tttt
import os
import gradio as gr
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from transformers import StoppingCriteria, StoppingCriteriaList, TextIteratorStreamer
from transformers import is_torch_npu_available
from threading import Thread


tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen1.5-7B-Chat")
model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen1.5-7B-Chat", torch_dtype=torch.bfloat16)
if is_torch_npu_available():
    model.to("npu:0")
elif torch.cuda.is_available():
    mode.to("cuda:0")


class StopOnTokens(StoppingCriteria):
    def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs) -> bool:
        stop_ids = [2]
        for stop_id in stop_ids:
            if input_ids[0][-1] == stop_id:
                return True
        return False


def predict(message, history):
    #if is_torch_npu_available():
    #    torch.npu.set_device(model.device)
    stop = StopOnTokens()
    conversation = []

    for user, assistant in history:
        conversation.extend([{"role": "user", "content": user}, {"role": "assistant", "content": assistant}])

    conversation.append({"role": "user", "content": message})
    print(f'>>>conversation={conversation}', flush=True)
    prompt = tokenizer.apply_chat_template(conversation, tokenize=False, add_generation_prompt=True)
    model_inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
    streamer = TextIteratorStreamer(tokenizer, timeout=100., skip_prompt=True, skip_special_tokens=True)
    generate_kwargs = dict(
        model_inputs,
        streamer=streamer,
        max_new_tokens=1024,
        do_sample=True,
        top_p=0.95,
        top_k=50,
        temperature=0.7,
        repetition_penalty=1.0,
        num_beams=1,
        stopping_criteria=StoppingCriteriaList([stop])
    )
    t = Thread(target=model.generate, kwargs=generate_kwargs)
    t.start()
    partial_message = ""
    for new_token in streamer:
        partial_message += new_token
        if '</s>' in partial_message:
            break
        yield partial_message


# Setting up the Gradio chat interface.
gr.ChatInterface(predict,
                 title="Qwen1.5 0.5B Chat Demo",
                 description="Warning. All answers are generated and may contain inaccurate information.",
                 examples=['How do you cook fish?', 'Who is the president of the United States?']
                 ).launch()