HeRksTAn commited on
Commit
041ea1b
1 Parent(s): 9a1d280
Files changed (2) hide show
  1. __pycache__/app.cpython-311.pyc +0 -0
  2. app.py +9 -4
__pycache__/app.cpython-311.pyc CHANGED
Binary files a/__pycache__/app.cpython-311.pyc and b/__pycache__/app.cpython-311.pyc differ
 
app.py CHANGED
@@ -12,6 +12,11 @@ from langchain.schema.runnable.config import RunnableConfig
12
  from langchain_core.output_parsers import StrOutputParser
13
  from langchain.text_splitter import RecursiveCharacterTextSplitter
14
  from langchain_community.document_loaders import UnstructuredPDFLoader
 
 
 
 
 
15
 
16
 
17
  load_dotenv()
@@ -77,10 +82,10 @@ text_splitter = RecursiveCharacterTextSplitter(
77
  # )
78
 
79
 
80
- # loader = UnstructuredPDFLoader("https://w2l.sbst.dk/file/502104/br_femogfirs.pdf")
81
- loader = UnstructuredPDFLoader("br_femogfirs.pdf")
82
- # data = loader.load_and_split(text_splitter)
83
- data = loader.load()
84
 
85
  # embedding_model = OpenAIEmbeddings(model="text-embedding-3-small")
86
 
 
12
  from langchain_core.output_parsers import StrOutputParser
13
  from langchain.text_splitter import RecursiveCharacterTextSplitter
14
  from langchain_community.document_loaders import UnstructuredPDFLoader
15
+ from langchain_community.document_loaders import OnlinePDFLoader
16
+ from langchain_community.document_loaders import PyPDFLoader
17
+
18
+
19
+
20
 
21
 
22
  load_dotenv()
 
82
  # )
83
 
84
 
85
+ # loader = PyPDFLoader("https://w2l.sbst.dk/file/502104/br_femogfirs.pdf")
86
+ loader = UnstructuredPDFLoader("br_femogfirs.pdf", strategy="fast")
87
+ data = loader.load_and_split(text_splitter)
88
+ # data = loader.load()
89
 
90
  # embedding_model = OpenAIEmbeddings(model="text-embedding-3-small")
91