indikamk commited on
Commit
547cc67
1 Parent(s): 7c6a0dd

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +27 -0
app.py CHANGED
@@ -1,3 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  import gradio as gr
2
 
3
  def greet(name):
 
1
+ from langchain.document_loaders import BSHTMLLoader, DirectoryLoader
2
+
3
+ bshtml_dir_loader = DirectoryLoader('./data/', loader_cls=BSHTMLLoader)
4
+
5
+ data = bshtml_dir_loader.load()
6
+
7
+ from langchain.text_splitter import RecursiveCharacterTextSplitter
8
+
9
+ text_splitter = RecursiveCharacterTextSplitter(
10
+ chunk_size = 1000,
11
+ chunk_overlap = 20,
12
+ length_function = len,
13
+ )
14
+
15
+ documents = text_splitter.split_documents(data)
16
+
17
+ import os
18
+
19
+ os.environ["OPENAI_API_KEY"] = "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"
20
+
21
+ from langchain.embeddings.openai import OpenAIEmbeddings
22
+
23
+ embeddings = OpenAIEmbeddings()
24
+
25
+
26
+
27
+
28
  import gradio as gr
29
 
30
  def greet(name):