Spaces:
Sleeping
Sleeping
TusharsinghBaghel
commited on
Commit
•
befa928
1
Parent(s):
503e330
Upload 6 files
Browse files- .env +5 -0
- Faiss_index/index.faiss +0 -0
- Faiss_index/index.pkl +3 -0
- __pycache__/main.cpython-39.pyc +0 -0
- main.py +90 -0
- requirements.txt +21 -0
.env
ADDED
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
'''
|
2 |
+
LANGCHAIN_API_KEY="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"
|
3 |
+
GROQ_API_KEY="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"
|
4 |
+
LANGCHAIN_PROJECT="try"
|
5 |
+
'''
|
Faiss_index/index.faiss
ADDED
Binary file (903 kB). View file
|
|
Faiss_index/index.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bd5de032dd6d0e0f86fddbe8b2feae0c725f9b4d73c788e94a10789adea59205
|
3 |
+
size 813439
|
__pycache__/main.cpython-39.pyc
ADDED
Binary file (2.97 kB). View file
|
|
main.py
ADDED
@@ -0,0 +1,90 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
import os
|
3 |
+
os.environ["KMP_DUPLICATE_LIB_OK"]="TRUE"
|
4 |
+
|
5 |
+
import warnings
|
6 |
+
from langchain_groq import ChatGroq
|
7 |
+
from langchain_community.embeddings import OllamaEmbeddings
|
8 |
+
from langchain_community.embeddings import HuggingFaceEmbeddings
|
9 |
+
|
10 |
+
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
11 |
+
|
12 |
+
from langchain.chains.combine_documents import create_stuff_documents_chain
|
13 |
+
from langchain_core.prompts import ChatPromptTemplate
|
14 |
+
from langchain.chains import create_retrieval_chain
|
15 |
+
from langchain_community.vectorstores import FAISS
|
16 |
+
from langchain_community.document_loaders import PyPDFDirectoryLoader
|
17 |
+
from typing import Any
|
18 |
+
from langchain.agents import AgentExecutor
|
19 |
+
from langchain_core.tools import tool
|
20 |
+
|
21 |
+
from dotenv import load_dotenv
|
22 |
+
from fastapi.middleware.cors import CORSMiddleware
|
23 |
+
|
24 |
+
load_dotenv()
|
25 |
+
groq_api_key=os.getenv('GROQ_API_KEY')
|
26 |
+
|
27 |
+
|
28 |
+
warnings.filterwarnings('ignore')
|
29 |
+
|
30 |
+
llm=ChatGroq(groq_api_key=groq_api_key,
|
31 |
+
model_name="mixtral-8x7b-32768")
|
32 |
+
|
33 |
+
prompt=ChatPromptTemplate.from_template(
|
34 |
+
"""
|
35 |
+
You are a Mineral Exploration assistant.You are given a context and an input query. Use the context to answer the query.
|
36 |
+
<context>
|
37 |
+
{context}
|
38 |
+
<context>
|
39 |
+
Questions:{input}
|
40 |
+
|
41 |
+
"""
|
42 |
+
)
|
43 |
+
|
44 |
+
def vector_embedding():
|
45 |
+
|
46 |
+
if "vectors" not in st.session_state:
|
47 |
+
|
48 |
+
|
49 |
+
print("Embeddings defined")
|
50 |
+
loader=PyPDFDirectoryLoader("./rules_regulations") ## Data Ingestion
|
51 |
+
docs=loader.load() ## Document Loading
|
52 |
+
print("PDF loaded")
|
53 |
+
text_splitter=RecursiveCharacterTextSplitter(chunk_size=4000,chunk_overlap=500) ## Chunk Creation
|
54 |
+
print("Chunks created")
|
55 |
+
final_documents=text_splitter.split_documents(docs) #splitting
|
56 |
+
print("Splitting done")
|
57 |
+
vectors=FAISS.from_documents(final_documents,embeddings) #vector embeddings
|
58 |
+
vectors.save_local("Faiss_index")
|
59 |
+
print("Embeddings created")
|
60 |
+
else:
|
61 |
+
print("Faiss index already exist")
|
62 |
+
|
63 |
+
|
64 |
+
|
65 |
+
|
66 |
+
|
67 |
+
embeddings=HuggingFaceEmbeddings()
|
68 |
+
vectors = FAISS.load_local("Faiss_index", embeddings, allow_dangerous_deserialization=True)
|
69 |
+
|
70 |
+
def Get_Rag_Response(query):
|
71 |
+
if query:
|
72 |
+
print("Got query:")
|
73 |
+
print(query)
|
74 |
+
document_chain=create_stuff_documents_chain(llm,prompt)
|
75 |
+
retriever=vectors.as_retriever()
|
76 |
+
print("Retrieved docs")
|
77 |
+
retrieval_chain=create_retrieval_chain(retriever,document_chain)
|
78 |
+
response=retrieval_chain.invoke({'input':query})
|
79 |
+
return response
|
80 |
+
|
81 |
+
else:
|
82 |
+
print("Error: No query passed")
|
83 |
+
|
84 |
+
iface= gr.Interface(fn = Get_Rag_Response,
|
85 |
+
inputs = ["text"],
|
86 |
+
outputs = ['text'],
|
87 |
+
title = "Rag_Response",
|
88 |
+
description="Get RAG chatbot response")
|
89 |
+
|
90 |
+
iface.launch(inline=False)
|
requirements.txt
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
langchain_core
|
2 |
+
python-dotenv
|
3 |
+
streamlit
|
4 |
+
langchain_community
|
5 |
+
sse_starlette
|
6 |
+
bs4
|
7 |
+
pypdf
|
8 |
+
faiss-cpu
|
9 |
+
groq
|
10 |
+
cassio
|
11 |
+
beautifulsoup4
|
12 |
+
langchain-groq
|
13 |
+
langchainhub
|
14 |
+
langchain
|
15 |
+
sentence_transformers
|
16 |
+
sentence-transformers
|
17 |
+
InstructorEmbedding
|
18 |
+
torch
|
19 |
+
urllib3<2
|
20 |
+
requests
|
21 |
+
gradio
|