File size: 2,139 Bytes
88d1ce9
 
4b8b036
88d1ce9
4b8b036
88d1ce9
 
 
 
4962f60
88d1ce9
 
 
 
 
 
 
 
4b8b036
88d1ce9
 
 
 
 
 
 
 
4b8b036
 
 
81c47bb
 
 
 
4b8b036
88d1ce9
4b8b036
 
88d1ce9
4b8b036
 
 
 
 
 
 
 
 
 
 
88d1ce9
 
 
 
 
 
 
3a06e2d
88d1ce9
 
 
 
 
 
 
 
 
 
 
4b8b036
88d1ce9
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
import gradio as gr
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
import spaces
import threading

model_name = "baidu/ERNIE-4.5-21B-A3B-Thinking"

# Load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    device_map="auto",
    torch_dtype=torch.bfloat16,
)

@spaces.GPU(duration=120)
def chat(message, history):
    messages = history + [{"role": "user", "content": message}]
    
    text = tokenizer.apply_chat_template(
        messages,
        tokenize=False,
        add_generation_prompt=True
    )
    model_inputs = tokenizer([text], add_special_tokens=False, return_tensors="pt").to(model.device)
    
    streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
    generation_kwargs = {
        **model_inputs,
        "streamer": streamer,
        "max_new_tokens": 1024,
        "do_sample": True,
        "temperature": 0.7,
    }
    
    thread = threading.Thread(target=model.generate, kwargs=generation_kwargs)
    thread.start()
    
    new_history = history + [{"role": "user", "content": message}]
    new_history.append({"role": "assistant", "content": ""})
    yield new_history, ""
    
    generated_text = ""
    for new_token in streamer:
        generated_text += new_token
        new_history[-1]["content"] = generated_text
        yield new_history, ""
    
    thread.join()

with gr.Blocks(title="ERNIE Chat") as demo:
    gr.Markdown("# ERNIE-4.5-21B-A3B-Thinking Chat App")
    chatbot = gr.Chatbot(
        height=500,
        type="messages",
        show_copy_button=True,
        avatar_images=None
    )
    msg = gr.Textbox(
        placeholder="Type your message here...",
        show_label=False,
        container=True,
        scale=7,
    )
    with gr.Row():
        clear_btn = gr.Button("Clear", variant="secondary")
    
    msg.submit(chat, [msg, chatbot], [chatbot, msg])
    clear_btn.click(lambda: ([], ""), None, [chatbot, msg], queue=False)

if __name__ == "__main__":
    demo.launch()