Update app.py

app.py

@@ -9,6 +9,7 @@ import os
 os.environ['OPENAI_API_KEY'] = os.environ['OPEN_API_KEY']
 st.title("Contracts Multiple File Search ")
 import pandas as pd
+import time
 
 from langchain.retrievers import BM25Retriever, EnsembleRetriever
 from langchain.schema import Document
@@ -19,7 +20,7 @@ from FlagEmbedding import FlagReranker
 reranker = FlagReranker('BAAI/bge-reranker-base')
 import spacy
 # Load the English model from SpaCy
-nlp = spacy.load("en_core_web_md")
+nlp = spacy.load("en_core_web_sm")
 
 def util_upload_file_and_return_list_docs(uploaded_files):
     #util_del_cwd()
@@ -180,7 +181,7 @@ def util_openai_extract_clause(example_prompt, page_content):
     openai.api_key = os.environ['OPENAI_API_KEY']
     content = example_prompt
     content = content + "\n Answer precisely; do not add anything extra, and try to locate the answer in the below context   \n context: "
-    return_value = openai.ChatCompletion.create(model="gpt-3.5-turbo",temperature=0.0001,messages=[{"role": "user", "content": content + "\n" + page_content},])
+    return_value = openai.ChatCompletion.create(model="gpt-4o-mini",temperature=0.0001,messages=[{"role": "user", "content": content + "\n" + page_content},])
     return(str(return_value['choices'][0]['message']['content']))
 
 
@@ -188,7 +189,7 @@ def util_openai_hyde(example_prompt):
     import openai
     openai.api_key = os.environ['OPENAI_API_KEY']
     content = example_prompt
-    return_value = openai.ChatCompletion.create(model="gpt-3.5-turbo",temperature=0.0001,messages=[
+    return_value = openai.ChatCompletion.create(model="gpt-4o-mini",temperature=0.0001,messages=[
         {"role": "system", "content": "You are a legal contract lawyer. generate a summary from below text " + "\n"},
         {"role": "user", "content": example_prompt + "\n"},
     
@@ -216,7 +217,7 @@ def util_openai_modify_prompt(example_prompt, page_content):
     openai.api_key = os.environ['OPENAI_API_KEY']
     my_prompt = """Expand the original Query to show exact resuls for extraction\n
     Query: """ + example_prompt # + """\nDocument: """ + page_content + """ """
-    return_value = openai.ChatCompletion.create(model="gpt-4",temperature=0.0001,messages=[{"role": "user", "content": my_prompt},])
+    return_value = openai.ChatCompletion.create(model="gpt-4o-mini",temperature=0.0001,messages=[{"role": "user", "content": my_prompt},])
     return(str(return_value['choices'][0]['message']['content']))    
 
 # def create_bm25_page_rank(page_list_retrieve, page_query):
@@ -273,6 +274,7 @@ with st.form("my_form"):
         ensemble_retriever = EnsembleRetriever(retrievers=[bm25_retriever, chroma_retriever],weights=[0.25, 0.75])
         passage_nodes = ensemble_retriever.get_relevant_documents(single_example_passage)
         print('len(passage_nodes):', len(passage_nodes))
+        time.sleep(5)
         df = util_get_list_pageno_and_contents(single_example_passage,passage_documents,passage_nodes)
         st.write(df)
         # print('len(page_list_retrieve):', len(page_list_retrieve))