File size: 2,420 Bytes
e348efe
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9814f59
7432eb1
e348efe
 
 
 
1fed219
 
9814f59
 
c96ea95
 
 
 
 
01127eb
a03faf2
c96ea95
 
9814f59
 
 
e348efe
 
 
 
addace4
 
 
 
 
e844d1b
addace4
 
 
 
 
 
 
9814f59
addace4
 
e844d1b
 
 
 
 
 
 
addace4
e844d1b
 
c96ea95
4516329
b3d631d
c96ea95
 
addace4
9814f59
 
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
import gradio as gr

from langchain.document_loaders import OnlinePDFLoader

from langchain.text_splitter import CharacterTextSplitter
text_splitter = CharacterTextSplitter(chunk_size=350, chunk_overlap=0)

from langchain.llms import HuggingFaceHub
flan_ul2 = HuggingFaceHub(repo_id="google/flan-ul2", model_kwargs={"temperature":0.1, "max_new_tokens":300})

from langchain.embeddings import HuggingFaceHubEmbeddings
embeddings = HuggingFaceHubEmbeddings()

from langchain.vectorstores import Chroma

from langchain.chains import RetrievalQA

def pdf_changes(pdf_doc):
    loader = OnlinePDFLoader(pdf_doc.name)
    documents = loader.load()
    texts = text_splitter.split_documents(documents)
    db = Chroma.from_documents(texts, embeddings)
    retriever = db.as_retriever()
    global qa 
    qa = RetrievalQA.from_chain_type(llm=flan_ul2, chain_type="stuff", retriever=retriever, return_source_documents=True)
    return "Ready"

def add_text(history, text):
    history = history + [(text, None)]
    return history, ""

def bot(history):
    response = infer(history[-1][0])
    history[-1][1] = response['result']
    return history

def infer(question):
    
    query = question
    result = qa({"query": query})

    return result

css="""
#col-container {max-width: 700px; margin-left: auto; margin-right: auto;}
"""

title = """
<div style="text-align: center;max-width: 700px;">
    <h1>Chat with PDF</h1>
    <p style="text-align: center;">Upload a .PDF from your computer, click the "Load PDF to LangChain" button, <br />
    when everything is ready, you can start asking questions about the pdf ;)</p>
</div>
"""


with gr.Blocks() as demo:
    with gr.Column(elem_id="col-container"):
        gr.HTML(title)
        
        with gr.Row():
            pdf_doc = gr.File(label="Load a pdf", file_types=['.pdf'], type="file")
            langchain_status = gr.Textbox(label="Status")
            load_pdf = gr.Button("Load pdf to langchain")
        
        chatbot = gr.Chatbot([], elem_id="chatbot").style(height=350)
        with gr.Row():
            question = gr.Textbox(label="Question")
            clear = gr.Button("Clear")
        
    load_pdf.click(pdf_changes, pdf_doc, langchain_status, queue=False)
    question.submit(add_text, [chatbot, question], [chatbot, question]).then(
        bot, chatbot, chatbot
    )
    clear.click(lambda: None, None, chatbot, queue=False)

demo.launch()