File size: 2,707 Bytes
c2c5723
 
 
cc5b602
c2c5723
 
 
 
 
 
51a7d9e
6386510
c2c5723
 
6386510
c2c5723
da59244
c2c5723
 
b70c257
c2c5723
 
 
 
 
 
652620b
c2c5723
7cb9567
c2c5723
 
 
 
 
 
 
 
b70c257
c2c5723
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b70c257
c2c5723
 
 
 
 
 
 
 
b70c257
c2c5723
 
 
 
 
f77fb99
c2c5723
 
b70c257
c2c5723
 
51a7d9e
c2c5723
51a7d9e
c2c5723
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
51a7d9e
c2c5723
51a7d9e
c2c5723
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
from langchain_community.vectorstores import Qdrant
from langchain_groq import ChatGroq
from langchain_huggingface import HuggingFaceEmbeddings
import os
from dotenv import load_dotenv
from langchain.prompts import ChatPromptTemplate
from langchain.schema.runnable import RunnablePassthrough
from langchain.schema.output_parser import StrOutputParser
from qdrant_client import QdrantClient, models
from langchain_qdrant import Qdrant
import gradio as gr

# Load environment variables
load_dotenv()

os.environ["GROQ_API_KEY"] = os.getenv("GROQ_API")

# HuggingFace Embeddings
embeddings = HuggingFaceEmbeddings(model_name="BAAI/bge-large-en-v1.5")

# Qdrant Client Setup
client = QdrantClient(
    url=os.getenv("QDRANT_URL"),
    api_key=os.getenv("QDRANT_API_KEY"),
    prefer_grpc=True
)

collection_name = "mawared"

# Try to create collection, handle if it already exists
try:
    client.create_collection(
        collection_name=collection_name,
        vectors_config=models.VectorParams(
            size=768,  # GTE-large embedding size
            distance=models.Distance.COSINE
        ),
    )
    print(f"Created new collection: {collection_name}")
except Exception as e:
    if "already exists" in str(e):
        print(f"Collection {collection_name} already exists, continuing...")
    else:
        raise e

# Create Qdrant vector store
db = Qdrant(
    client=client,
    collection_name=collection_name,
    embeddings=embeddings,
)

# Create retriever
retriever = db.as_retriever(
    search_type="similarity",
    search_kwargs={"k": 5}
)

# LLM setup
llm = ChatGroq(
    model="llama-3.3-70b-versatile",
    temperature=0.1,
    max_tokens=None,
    timeout=None,
    max_retries=2,
)

# Create prompt template
template = """
You are an expert assistant specializing in the LONG COT RAG. Your task is to answer the user's question strictly based on the provided context...
Context:
{context}

Question:
{question}

Answer:
"""

prompt = ChatPromptTemplate.from_template(template)

# Create the RAG chain
rag_chain = (
    {"context": retriever, "question": RunnablePassthrough()}
    | prompt
    | llm
    | StrOutputParser()
)

# Define the Gradio function
def ask_question_gradio(question):
    result = ""
    for chunk in rag_chain.stream(question):
        result += chunk
    return result

# Create the Gradio interface
interface = gr.Interface(
    fn=ask_question_gradio,
    inputs="text",
    outputs="text",
    title="Mawared Expert Assistant",
    description="Ask questions about the Mawared HR System or any related topic using Chain-of-Thought (CoT) and RAG principles.",
    theme="compact",
)

# Launch Gradio app
if __name__ == "__main__":
    interface.launch()