File size: 4,103 Bytes
4602937
fada25c
4615482
4602937
fada25c
 
4602937
2b44908
 
fada25c
2b44908
fada25c
 
2b44908
 
fada25c
 
 
 
 
 
 
 
3430157
fada25c
 
 
 
2b44908
fada25c
 
 
2b44908
 
 
fada25c
 
 
 
 
 
 
6dd9499
fada25c
 
 
6dd9499
 
 
 
 
 
fada25c
 
 
 
 
 
 
 
6dd9499
 
 
 
 
 
 
 
fada25c
 
2b44908
fada25c
2b44908
fada25c
2b44908
fada25c
6dd9499
 
fada25c
2b44908
 
 
fada25c
 
 
6dd9499
ddc78b4
6dd9499
 
fada25c
5f6f331
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
138
139
140
141
from dotenv import load_dotenv
import gradio as gr
import os
from llama_index.core import StorageContext, load_index_from_storage, VectorStoreIndex, SimpleDirectoryReader, ChatPromptTemplate, Settings
from llama_index.llms.huggingface import HuggingFaceInferenceAPI
from llama_index.embeddings.huggingface import HuggingFaceEmbedding
from sentence_transformers import SentenceTransformer

# Load environment variables
load_dotenv()

# Configure the Llama index settings
Settings.llm = HuggingFaceInferenceAPI(
    model_name="meta-llama/Meta-Llama-3-8B-Instruct",
    tokenizer_name="meta-llama/Meta-Llama-3-8B-Instruct",
    context_window=3000,
    token=os.getenv("HF_TOKEN"),
    max_new_tokens=512,
    generate_kwargs={"temperature": 0.1},
)
Settings.embed_model = HuggingFaceEmbedding(
    model_name="BAAI/bge-small-en-v1.5"
)

# Define the directory for persistent storage and data
PERSIST_DIR = "db"
PDF_DIRECTORY = 'data'  # Changed to the directory containing PDFs

# Ensure directories exist
os.makedirs(PDF_DIRECTORY, exist_ok=True)
os.makedirs(PERSIST_DIR, exist_ok=True)

# Variable to store current chat conversation
current_chat_history = []

def data_ingestion_from_directory():
    # Use SimpleDirectoryReader on the directory containing the PDF files
    documents = SimpleDirectoryReader(PDF_DIRECTORY).load_data()
    storage_context = StorageContext.from_defaults()
    index = VectorStoreIndex.from_documents(documents)
    index.storage_context.persist(persist_dir=PERSIST_DIR)

def handle_query(query):
    chat_text_qa_msgs = [
        (
            "user",
            """
            You are now the RedFerns Tech chatbot. Your aim is to provide answers to the user based on the conversation flow only.
            {context_str}
            Question:
            {query_str}
            """
        )
    ]
    text_qa_template = ChatPromptTemplate.from_messages(chat_text_qa_msgs)

    # Load index from storage
    storage_context = StorageContext.from_defaults(persist_dir=PERSIST_DIR)
    index = load_index_from_storage(storage_context)

    # Use chat history to enhance response
    context_str = ""
    for past_query, response in reversed(current_chat_history):
        if past_query.strip():
            context_str += f"User asked: '{past_query}'\nBot answered: '{response}'\n"

    query_engine = index.as_query_engine(text_qa_template=text_qa_template, context_str=context_str)
    answer = query_engine.query(query)

    if hasattr(answer, 'response'):
        response = answer.response
    elif isinstance(answer, dict) and 'response' in answer:
        response = answer['response']
    else:
        response = "Sorry, I couldn't find an answer."

    # Update current chat history
    current_chat_history.append((query, response))

    return response

# Example usage: Process PDF ingestion from directory
print("Processing PDF ingestion from directory:", PDF_DIRECTORY)
data_ingestion_from_directory()

# Define the function to handle predictions
def predict(message,history):
    response = handle_query(message)
    return response


def custom_layout(component):
  css = '''
  /* Style the chat container */
  .gradio-container {
    display: flex;
    flex-direction: column;
    width: 400px;
    margin: 0 auto;
    padding: 20px;
    border: 1px solid #ddd;
    border-radius: 5px;
    background-color: #eee;
  }

  /* Style the title */
  .gradio-title {
    text-align: center;
    font-weight: bold;
    margin-bottom: 10px;
  }

  /* Style the chat history */
  .gradio-chat-history {
    flex: 1;
    overflow-y: scroll;
    padding: 10px;
    border-bottom: 1px solid #ddd;
  }

  /* Style the chat message */
  .gradio-message {
    margin-bottom: 5px;
  }

  /* Style the user input field */
  .gradio-chat-input input {
    width: 100%;
    padding: 10px;
    border: 1px solid #ccc;
    border-radius: 3px;
    font-size: 16px;
  }
  '''
  return gr.components.HTML(css)(component)

# Launch the chat interface with custom layout
interface = gr.ChatInterface(predict, title="RedFerns Tech", layout=custom_layout)
interface.launch()