File size: 3,994 Bytes
b4b5bdf
e9698e9
e0e448c
e9698e9
 
 
 
3fcc7da
 
b4b5bdf
51727c4
 
 
 
 
 
e9698e9
 
 
 
e0e448c
e9698e9
1203b67
e9698e9
 
 
 
 
 
 
 
 
 
2785052
 
69a190d
 
 
2785052
e0e448c
e9698e9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3fcc7da
 
 
 
 
 
 
 
 
e0e448c
e9698e9
 
3fcc7da
 
 
 
 
e9698e9
 
 
 
 
 
 
 
 
3fcc7da
e9698e9
 
 
 
 
 
 
3fcc7da
 
 
 
 
e0e448c
 
e9698e9
 
 
 
 
 
 
 
b4b5bdf
e9698e9
 
a3a378d
e9698e9
 
 
 
3fcc7da
 
e9698e9
 
 
 
 
e0e448c
e9698e9
 
e0e448c
e9698e9
 
 
 
7710388
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
136
137
import logging
import os
from typing import Optional

import gradio as gr
import pandas as pd

from buster.completers import Completion

import cfg
from cfg import setup_buster

buster = setup_buster(cfg.buster_cfg)

#  suppress httpx logs they are spammy and uninformative
logging.getLogger("httpx").setLevel(logging.WARNING)

logger = logging.getLogger(__name__)
logging.basicConfig(level=logging.INFO)

AVAILABLE_SOURCES = ["towardsai", "wikipedia", "langchain_course"]


def format_sources(matched_documents: pd.DataFrame) -> str:
    if len(matched_documents) == 0:
        return ""

    documents_answer_template: str = "πŸ“ Here are the sources I used to answer your question:\n\n{documents}\n\n{footnote}"
    document_template: str = "[πŸ”— {document.title}]({document.url}), relevance: {document.similarity_to_answer:2.1f} %"

    matched_documents.similarity_to_answer = (
        matched_documents.similarity_to_answer * 100
    )

    # drop duplicates, keep highest ranking ones
    matched_documents = matched_documents.sort_values(
        "similarity_to_answer", ascending=False
    ).drop_duplicates("title", keep="first")

    documents = "\n".join(
        [
            document_template.format(document=document)
            for _, document in matched_documents.iterrows()
        ]
    )
    footnote: str = "I'm a bot πŸ€– and not always perfect."

    return documents_answer_template.format(documents=documents, footnote=footnote)


def add_sources(history, completion):
    if completion.answer_relevant:
        formatted_sources = format_sources(completion.matched_documents)
        history.append([None, formatted_sources])

    return history


def user(user_input, history):
    """Adds user's question immediately to the chat."""
    return "", history + [[user_input, None]]


def get_empty_source_completion(user_input):
    return Completion(
        user_input=user_input,
        answer_text="You have to select at least one source from the dropdown menu.",
        matched_documents=pd.DataFrame(),
        error=False,
    )


def get_answer(history, sources: Optional[list[str]] = None):
    user_input = history[-1][0]

    if len(sources) == 0:
        completion = get_empty_source_completion(user_input)

    else:
        completion = buster.process_input(user_input, sources=sources)

    history[-1][1] = ""

    for token in completion.answer_generator:
        history[-1][1] += token

        yield history, completion


block = gr.Blocks()

with block:
    with gr.Row():
        gr.Markdown(
            "<h3><center>Buster πŸ€–: A Question-Answering Bot for your documentation</center></h3>"
        )

    source_selection = gr.Dropdown(
        choices=AVAILABLE_SOURCES,
        label="Select Sources",
        value=AVAILABLE_SOURCES,
        multiselect=True,
    )

    chatbot = gr.Chatbot()

    with gr.Row():
        question = gr.Textbox(
            label="What's your question?",
            placeholder="Ask a question to AI stackoverflow here...",
            lines=1,
        )
        submit = gr.Button(value="Send", variant="secondary")

    examples = gr.Examples(
        examples=cfg.example_questions,
        inputs=question,
    )

    gr.Markdown(
        "This application uses ChatGPT to search the docs for relevant info and answer questions. "
        "\n\n### Powered by [Buster πŸ€–](www.github.com/jerpint/buster)"
    )

    response = gr.State()

    submit.click(user, [question, chatbot], [question, chatbot], queue=False).then(
        get_answer, inputs=[chatbot, source_selection], outputs=[chatbot, response]
    ).then(add_sources, inputs=[chatbot, response], outputs=[chatbot])
    question.submit(user, [question, chatbot], [question, chatbot], queue=False).then(
        get_answer, inputs=[chatbot, source_selection], outputs=[chatbot, response]
    ).then(add_sources, inputs=[chatbot, response], outputs=[chatbot])


block.queue(concurrency_count=16)
block.launch(debug=True, share=False)