File size: 5,411 Bytes
aadc8fb
 
 
 
 
 
 
 
 
5edaa8a
aadc8fb
 
 
 
afe5a38
 
 
304ee09
aadc8fb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
74
75
76
77
78
79
80
81
import spaces
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
import os
import gradio as gr
import sentencepiece


os.environ['PYTORCH_CUDA_ALLOC_CONF'] = 'max_split_size_mb:120'
model_id = "01-ai/Yi-1.5-9B-Chat"
tokenizer_path = "./"
# eos_token_id = 7

DESCRIPTION = """
# 欢迎来到 Tonic的 YI-9B基地!👋🏻Welcome to 🙋🏻‍♂️Tonic's🧑🏻‍🚀Yi-1.5-9B-Chat!🚀
You can use this Space to test out the current model 您可以使用此空间测试当前模型 [01-ai/Yi-9B](https://huggingface.co/01-ai/Yi-1.5-9B-Chat)  您也可以通过克隆这个空间来使用 YI-9B基地 只需点击这里".
You can also use 🧑🏻‍🚀01-ai/Yi-1.5-9B-Chat🚀 by cloning this space. 🧬🔬🔍 Simply click here: <a style="display:inline-block" href="https://huggingface.co/spaces/Tonic1/Yi-9B?duplicate=true"><img src="https://img.shields.io/badge/-Duplicate%20Space-blue?labelColor=white&style=flat&logo=data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAAAXNSR0IArs4c6QAAAP5JREFUOE+lk7FqAkEURY+ltunEgFXS2sZGIbXfEPdLlnxJyDdYB62sbbUKpLbVNhyYFzbrrA74YJlh9r079973psed0cvUD4A+4HoCjsA85X0Dfn/RBLBgBDxnQPfAEJgBY+A9gALA4tcbamSzS4xq4FOQAJgCDwV2CPKV8tZAJcAjMMkUe1vX+U+SMhfAJEHasQIWmXNN3abzDwHUrgcRGmYcgKe0bxrblHEB4E/pndMazNpSZGcsZdBlYJcEL9Afo75molJyM2FxmPgmgPqlWNLGfwZGG6UiyEvLzHYDmoPkDDiNm9JR9uboiONcBXrpY1qmgs21x1QwyZcpvxt9NS09PlsPAAAAAElFTkSuQmCC&logoWidth=14" alt="Duplicate Space"></a></h3> 
加入我们:🌟TeamTonic 总是在制作酷炫的演示!加入我们活跃的建设者🛠️社区,在👻DDiscord](https://discord.gg/nXx5wbX9),在🤗Huggingface:[TeamTonic](https://huggingface.co/TeamTonic) & [MultiTransformer](https://huggingface.co/MultiTransformer) 在  在🌐Github [Tonic-AI](https://github.com/tonic-ai)上,为🌟[Multitonic](https://github.com/tonic-ai/multitonic)做出贡献。 Join us : 🌟TeamTonic🌟 is always making cool demos! Join our active builder's🛠️community on 👻Discord: [Discord](https://discord.gg/nXx5wbX9) On 🤗Huggingface: [TeamTonic](https://huggingface.co/TeamTonic) & [MultiTransformer](https://huggingface.co/MultiTransformer) On 🌐Github: [Tonic-AI](https://github.com/tonic-ai) & contribute to 🌟 [Multitonic](https://github.com/tonic-ai/multitonic)
"""

tokenizer = AutoTokenizer.from_pretrained(model_id, device_map="auto", trust_remote_code=True)
# tokenizer = YiTokenizer.from_pretrained(tokenizer_path)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", torch_dtype=torch.bfloat16, trust_remote_code=True)
# tokenizer.eos_token_id = eos_token_id
# model.config.eos_token_id = eos_token_id

def format_prompt(user_message, system_message="You are YiTonic, an AI language model created by Tonic-AI. You are a cautious assistant. You carefully follow instructions. You are helpful and harmless and follow ethical guidelines and promote positive behavior."):
    prompt = f"<|im_start|>assistant\n{system_message}<|im_end|>\n<|im_start|>\nuser\n{user_message}<|im_end|>\nassistant\n"
    return prompt

@spaces.GPU
def predict(message, system_message, max_new_tokens=600, temperature=3.5, top_p=0.9, top_k=40, do_sample=False):
    formatted_prompt = format_prompt(message, system_message)

    input_ids = tokenizer.encode(formatted_prompt, return_tensors='pt')
    input_ids = input_ids.to(model.device)

    response_ids = model.generate(
        input_ids,
        max_length=max_new_tokens + input_ids.shape[1],
        temperature=temperature,  
        top_p=top_p,              
        top_k=top_k,              
        no_repeat_ngram_size=9,
        pad_token_id=tokenizer.eos_token_id,
        do_sample=do_sample
    )

    response = tokenizer.decode(response_ids[:, input_ids.shape[-1]:][0], skip_special_tokens=True)
    truncate_str = "<|im_end|>"
    if truncate_str and truncate_str in response:
        response = response.split(truncate_str)[0]

    return [("bot", response)]    
    
with gr.Blocks() as demo:
    gr.Markdown(DESCRIPTION)
    with gr.Group():
        textbox = gr.Textbox(placeholder='Your Message Here', label='Your Message', lines=2)
        system_prompt = gr.Textbox(placeholder='Provide a System Prompt In The First Person', label='System Prompt', lines=2, value="You are YiTonic, an AI language model created by Tonic-AI. You are a cautious assistant. You carefully follow instructions. You are helpful and harmless and you follow ethical guidelines and promote positive behavior.")

    with gr.Group():
        chatbot = gr.Chatbot(label='TonicYi-9B-Base-🧠🤯')

    with gr.Group():
        submit_button = gr.Button('Submit', variant='primary')

    with gr.Accordion(label='Advanced options', open=False):
        max_new_tokens = gr.Slider(label='Max New Tokens', minimum=1, maximum=55000, step=1, value=4056)
        temperature = gr.Slider(label='Temperature', minimum=0.1, maximum=4.0, step=0.1, value=1.2)
        top_p = gr.Slider(label='Top-P (nucleus sampling)', minimum=0.05, maximum=1.0, step=0.05, value=0.9)
        top_k = gr.Slider(label='Top-K', minimum=1, maximum=1000, step=1, value=40)
        do_sample_checkbox = gr.Checkbox(label='Disable for faster inference', value=True)

    submit_button.click(
        fn=predict,
        inputs=[textbox, system_prompt, max_new_tokens, temperature, top_p, top_k, do_sample_checkbox],
        outputs=chatbot
    )

demo.launch()