Spaces:
Sleeping
Sleeping
angelesteban00
commited on
Commit
•
8b091a4
1
Parent(s):
168d589
- app.py +2 -2
- load_data_from_PDF.py +34 -0
- requirements.txt +2 -0
app.py
CHANGED
@@ -21,11 +21,11 @@ Demo based on https://www.mongodb.com/developer/products/atlas/rag-atlas-vector-
|
|
21 |
## Prerequisites:
|
22 |
create a free DB called "langchain_demo" and a collection called "collection_of_text_blobs" in MongoDB Atlas (https://cloud.mongodb.com). After that, you have two options:
|
23 |
|
24 |
-
**option1**) execute locally "load_data.py" to create new documents and their embeddings in MongoDB<br>
|
25 |
**option2**) import the JSON file "langchain_demo.collection_of_text_blobs.json"
|
26 |
|
27 |
## Dataset
|
28 |
-
The JSON documents in MongoDB looks like:
|
29 |
```
|
30 |
{
|
31 |
"_id": {
|
|
|
21 |
## Prerequisites:
|
22 |
create a free DB called "langchain_demo" and a collection called "collection_of_text_blobs" in MongoDB Atlas (https://cloud.mongodb.com). After that, you have two options:
|
23 |
|
24 |
+
**option1**) execute locally "load_data.py"/"load_data_from_PDF.py" to create new documents and their embeddings in MongoDB<br>
|
25 |
**option2**) import the JSON file "langchain_demo.collection_of_text_blobs.json"
|
26 |
|
27 |
## Dataset
|
28 |
+
The JSON documents in MongoDB looks like (also was splitted and embebed this PDF https://arxiv.org/pdf/2303.08774.pdf):
|
29 |
```
|
30 |
{
|
31 |
"_id": {
|
load_data_from_PDF.py
ADDED
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from pymongo import MongoClient
|
2 |
+
# error since Jan 2024, from langchain.embeddings.openai import OpenAIEmbeddings
|
3 |
+
from langchain_openai import OpenAIEmbeddings
|
4 |
+
# error since Jan 2024, from langchain.vectorstores import MongoDBAtlasVectorSearch
|
5 |
+
from langchain_community.vectorstores import MongoDBAtlasVectorSearch
|
6 |
+
# error since Jan 2024, from langchain.document_loaders import PyPDFLoader
|
7 |
+
from langchain_community.document_loaders import PyPDFLoader
|
8 |
+
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
9 |
+
import os
|
10 |
+
|
11 |
+
mongo_uri = os.getenv("MONGO_URI")
|
12 |
+
openai_api_key = os.getenv("OPENAI_API_KEY")
|
13 |
+
|
14 |
+
client = MongoClient(mongo_uri)
|
15 |
+
dbName = "langchain_demo"
|
16 |
+
collectionName = "collection_of_text_blobs"
|
17 |
+
collection = client[dbName][collectionName]
|
18 |
+
|
19 |
+
#loader = DirectoryLoader( './sample_files', glob="./*.txt", show_progress=True)
|
20 |
+
loader = PyPDFLoader("https://arxiv.org/pdf/2303.08774.pdf")
|
21 |
+
data = loader.load()
|
22 |
+
text_splitter = RecursiveCharacterTextSplitter(chunk_size = 500, chunk_overlap = 0)
|
23 |
+
docs = text_splitter.split_documents(data)
|
24 |
+
|
25 |
+
#embeddings = OpenAIEmbeddings(openai_api_key=openai_api_key)
|
26 |
+
#vectorStore = MongoDBAtlasVectorSearch.from_documents( data, embeddings, collection=collection, index_name="default" )
|
27 |
+
|
28 |
+
# insert the documents in MongoDB Atlas Vector Search
|
29 |
+
x = MongoDBAtlasVectorSearch.from_documents(
|
30 |
+
documents=docs,
|
31 |
+
embedding=OpenAIEmbeddings(openai_api_key=openai_api_key, disallowed_special=()),
|
32 |
+
collection=collection,
|
33 |
+
index_name="default"
|
34 |
+
)
|
requirements.txt
CHANGED
@@ -1,4 +1,6 @@
|
|
1 |
langchain
|
|
|
|
|
2 |
langchain-openai
|
3 |
pymongo[srv]==4.1.1
|
4 |
bs4
|
|
|
1 |
langchain
|
2 |
+
pypdf
|
3 |
+
python-dotenv
|
4 |
langchain-openai
|
5 |
pymongo[srv]==4.1.1
|
6 |
bs4
|