File size: 2,866 Bytes
7efd637
b6ef90b
 
c33dbd2
ca8dc25
5ee7ec4
b6ef90b
 
ca8dc25
c33dbd2
 
e98c6cb
c33dbd2
 
 
d107cdf
5ee7ec4
 
 
 
b6ef90b
 
 
 
9dc7fb7
b6ef90b
 
 
 
 
 
 
 
 
5ee7ec4
b6ef90b
 
 
 
 
 
 
 
6a8b740
c33dbd2
c0f1215
c33dbd2
b6ef90b
c33dbd2
 
6719d1c
b6ef90b
c33dbd2
b6ef90b
 
 
 
82ee039
b6ef90b
 
c0f1215
b6ef90b
 
 
 
 
 
 
 
 
 
 
c0f1215
b6ef90b
 
 
 
7efd637
c33dbd2
b6ef90b
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
import gradio as gr
from PIL import Image
import requests
import os
from together import Together
import base64
from threading import Thread
import time

# Initialize Together client
client = Together()

# Ensure API key is set
if "TOGETHER_API_KEY" not in os.environ:
    raise ValueError("Please set the TOGETHER_API_KEY environment variable")

def encode_image(image_path):
    with open(image_path, "rb") as image_file:
        return base64.b64encode(image_file.read()).decode('utf-8')

def bot_streaming(message, history, max_new_tokens=250):
    txt = message["text"]
    messages = []
    images = []

    for i, msg in enumerate(history):
        if isinstance(msg[0], tuple):
            messages.append({"role": "user", "content": [{"type": "text", "text": history[i+1][0]}, {"type": "image_url", "image_url": {"url": f"data:image/jpeg;base64,{encode_image(msg[0][0])}"}}]})
            messages.append({"role": "assistant", "content": [{"type": "text", "text": history[i+1][1]}]})
        elif isinstance(history[i-1], tuple) and isinstance(msg[0], str):
            pass
        elif isinstance(history[i-1][0], str) and isinstance(msg[0], str):
            messages.append({"role": "user", "content": [{"type": "text", "text": msg[0]}]})
            messages.append({"role": "assistant", "content": [{"type": "text", "text": msg[1]}]})

    if len(message["files"]) == 1:
        if isinstance(message["files"][0], str):  # examples
            image_path = message["files"][0]
        else:  # regular input
            image_path = message["files"][0]["path"]
        messages.append({"role": "user", "content": [{"type": "text", "text": txt}, {"type": "image_url", "image_url": {"url": f"data:image/jpeg;base64,{encode_image(image_path)}"}}]})
    else:
        messages.append({"role": "user", "content": [{"type": "text", "text": txt}]})

    stream = client.chat.completions.create(
        model="meta-llama/Llama-Vision-Free",
        messages=messages,
        max_tokens=max_new_tokens,
        stream=True,
    )

    buffer = ""
    for chunk in stream:
        if chunk.choices[0].delta.content is not None:
            buffer += chunk.choices[0].delta.content
            time.sleep(0.01)
            yield buffer

demo = gr.ChatInterface(
    fn=bot_streaming,
    title="Meta Llama 3.2 Vision 11B",
    textbox=gr.MultimodalTextbox(),
    additional_inputs=[
        gr.Slider(
            minimum=10,
            maximum=500,
            value=250,
            step=10,
            label="Maximum number of new tokens to generate",
        )
    ],
    cache_examples=False,
    description="Try Multimodal Llama by Meta with the Together API in this demo. Upload an image, and start chatting about it",
    stop_btn="Stop Generation",
    fill_height=True,
    multimodal=True
)

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