File size: 6,248 Bytes
7119802
 
3566a6b
983678f
 
cd7c7be
338a3b4
 
 
 
 
7119802
69601bd
 
 
3566a6b
338a3b4
266f102
7119802
983678f
 
 
 
 
 
 
 
 
 
 
 
 
338a3b4
 
 
 
 
 
 
 
4c0c0ee
338a3b4
7119802
9b0f500
 
d0790bd
 
 
 
 
9b0f500
7119802
df3bc75
d0790bd
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cd7c7be
d0790bd
df3bc75
d0790bd
 
6cc6de1
 
fa76721
 
 
 
 
 
 
 
 
0aa18a5
 
6cc6de1
 
 
d0790bd
df3bc75
 
d0790bd
4e35ae8
d0790bd
 
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
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
from huggingface_hub import InferenceClient
import gradio as gr
import os
import re
import requests
import random
import http.client
import typing
import urllib.request
import vertexai
from vertexai.generative_models import GenerativeModel, Image

with open(".config/application_default_credentials.json", 'w') as file:
    file.write(str(os.getenv('credentials')))

vertexai.init(project=os.getenv('project_id'))
model = GenerativeModel("gemini-1.0-pro-vision")
client = InferenceClient("google/gemma-7b-it")

def extract_image_urls(text):
    url_regex = r"(https?:\/\/.*\.(?:png|jpg|jpeg|gif|webp|svg))"
    image_urls = re.findall(url_regex, text, flags=re.IGNORECASE)
    valid_image_url = ""
    for url in image_urls:
        try:
            response = requests.head(url)  # Use HEAD request for efficiency
            if response.status_code in range(200, 300) and 'image' in response.headers.get('content-type', ''):
                valid_image_url = url
        except requests.exceptions.RequestException:
            pass  # Ignore inaccessible URLs
    return valid_image_url

def load_image_from_url(image_url: str) -> Image:
    with urllib.request.urlopen(image_url) as response:
        response = typing.cast(http.client.HTTPResponse, response)
        image_bytes = response.read()
    return Image.from_bytes(image_bytes)

def search(url):
    image = load_image_from_url(url)
    response = model.generate_content([image,"Describe what is shown in this image."])
    return response.text

def format_prompt(message, history, cust_p):
    prompt = ""
    if history:
        for user_prompt, bot_response in history:
            prompt += f"<start_of_turn>user{user_prompt}<end_of_turn>"
            prompt += f"<start_of_turn>model{bot_response}<end_of_turn>"
    prompt+=cust_p.replace("USER_INPUT",message)
    return prompt

def chat_inf(system_prompt,prompt,history,memory,seed,temp,tokens,top_p,rep_p,chat_mem,cust_p):
    hist_len=0
    if not history:
        history = []
        hist_len=0
    if not memory:
        memory = []
        mem_len=0        
    if memory:
        for ea in memory[0-chat_mem:]:
            hist_len+=len(str(ea))
    in_len=len(system_prompt+prompt)+hist_len

    if (in_len+tokens) > 8000:
        history.append((prompt,"Wait, that's too many tokens, please reduce the 'Chat Memory' value, or reduce the 'Max new tokens' value"))
        yield history,memory
    else:
        generate_kwargs = dict(
            temperature=temp,
            max_new_tokens=tokens,
            top_p=top_p,
            repetition_penalty=rep_p,
            do_sample=True,
            seed=seed,
        )
        
        image = extract_image_urls(prompt)
        if image:
            image_description = "Image Description: " + search(image)
            prompt = prompt.replace(image, image_description)
            print(prompt)
            
        if system_prompt:
            formatted_prompt = format_prompt(f"{system_prompt}, {prompt}", memory[0-chat_mem:],cust_p)
        else:
            formatted_prompt = format_prompt(prompt, memory[0-chat_mem:],cust_p)
        stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=True)
        output = ""
        for response in stream:
            output += response.token.text
            yield [(prompt,output)],memory
        history.append((prompt,output))
        memory.append((prompt,output))
        yield history,memory

def clear_fn():
    return None,None,None,None
rand_val=random.randint(1,1111111111111111)

def check_rand(inp,val):
    if inp==True:
        return gr.Slider(label="Seed", minimum=1, maximum=1111111111111111, value=random.randint(1,1111111111111111))
    else:
        return gr.Slider(label="Seed", minimum=1, maximum=1111111111111111, value=int(val))
    
with gr.Blocks(theme=gr.themes.Soft()) as app:
    memory=gr.State()
    gr.HTML("""<center><h1 style='font-size:xx-large;'>Gemma Gemini Multimodal Chatbot</h1><br><h3>Gemini Sprint submission by Rishiraj Acharya. Uses Google's Gemini 1.0 Pro Vision multimodal model from Vertex AI with Google's Gemma 7B Instruct model from Hugging Face. Google Cloud credits are provided for this project.</h3>""")
    chat_b = gr.Chatbot(show_label=True, show_share_button=True, show_copy_button=True, likeable=True, layout="bubble", bubble_full_width=False)
    with gr.Group():
        inp = gr.Textbox(label="User Prompt")
        sys_inp = gr.Textbox(label="System Prompt")
        with gr.Accordion("Settings",open=False):
            custom_prompt=gr.Textbox(label="Modify Prompt Format", info="For testing purposes. 'USER_INPUT' is where 'SYSTEM_PROMPT, PROMPT' will be placed", lines=3,value="<start_of_turn>userUSER_INPUT<end_of_turn><start_of_turn>model")
            rand = gr.Checkbox(label="Random Seed", value=True)
            seed=gr.Slider(label="Seed", minimum=1, maximum=1111111111111111,step=1, value=rand_val)
            tokens = gr.Slider(label="Max new tokens",value=1600,minimum=0,maximum=8000,step=64,interactive=True, visible=True,info="The maximum number of tokens")
            temp=gr.Slider(label="Temperature",step=0.01, minimum=0.01, maximum=1.0, value=0.49)
            top_p=gr.Slider(label="Top-P",step=0.01, minimum=0.01, maximum=1.0, value=0.49)
            rep_p=gr.Slider(label="Repetition Penalty",step=0.01, minimum=0.1, maximum=2.0, value=0.99)
            chat_mem=gr.Number(label="Chat Memory", info="Number of previous chats to retain",value=4)
    with gr.Group():
        with gr.Row():
            btn = gr.Button("Chat", variant="primary")
            stop_btn = gr.Button("Stop", variant="stop")
            clear_btn = gr.Button("Clear", variant="secondary")
    
    chat_sub=inp.submit(check_rand,[rand,seed],seed).then(chat_inf,[sys_inp,inp,chat_b,memory,seed,temp,tokens,top_p,rep_p,chat_mem,custom_prompt],[chat_b,memory])
    go=btn.click(check_rand,[rand,seed],seed).then(chat_inf,[sys_inp,inp,chat_b,memory,seed,temp,tokens,top_p,rep_p,chat_mem,custom_prompt],[chat_b,memory])
    
    stop_btn.click(None,None,None,cancels=[go,chat_sub])
    clear_btn.click(clear_fn,None,[inp,sys_inp,chat_b,memory])
app.queue(default_concurrency_limit=10).launch()