Merge pull request #74 from arc53/fix-dbqa

fix dbqa, with new chain type, also fix for doc export

application/app.py

@@ -5,8 +5,8 @@ import datetime
 from flask import Flask, request, render_template
 # os.environ["LANGCHAIN_HANDLER"] = "langchain"
 import faiss
-from langchain import OpenAI
-from langchain.chains import VectorDBQAWithSourcesChain
+from langchain import OpenAI, VectorDBQA
+from langchain.chains.question_answering import load_qa_chain
 from langchain.prompts import PromptTemplate
 import requests
 
@@ -69,11 +69,22 @@ def api_answer():
     c_prompt = PromptTemplate(input_variables=["summaries", "question"], template=template)
     # create a chain with the prompt template and the store
 
-    chain = VectorDBQAWithSourcesChain.from_llm(llm=OpenAI(openai_api_key=api_key, temperature=0), vectorstore=store, combine_prompt=c_prompt)
+    #chain = VectorDBQA.from_llm(llm=OpenAI(openai_api_key=api_key, temperature=0), vectorstore=store, combine_prompt=c_prompt)
+    # chain = VectorDBQA.from_chain_type(llm=OpenAI(openai_api_key=api_key, temperature=0), chain_type='map_reduce',
+    #                                    vectorstore=store)
+
+    qa_chain = load_qa_chain(OpenAI(openai_api_key=api_key, temperature=0), chain_type="map_reduce",
+                             combine_prompt=c_prompt)
+    chain = VectorDBQA(combine_documents_chain=qa_chain, vectorstore=store)
+
+
+
     # fetch the answer
-    result = chain({"question": question})
+    result = chain({"query": question})
+    print(result)
 
     # some formatting for the frontend
+    result['answer'] = result['result']
     result['answer'] = result['answer'].replace("\\n", "<br>")
     result['answer'] = result['answer'].replace("SOURCES:", "")
     # mock result

scripts/parser/open_ai_func.py

@@ -31,13 +31,20 @@ def call_openai_api(docs):
             print("Error on ", i)
             print("Saving progress")
             print(f"stopped at {c1} out of {len(docs)}")
-            store.save_local("outputs")
+            faiss.write_index(store.index, "docs.index")
+            store.index = None
+            with open("faiss_store.pkl", "wb") as f:
+                pickle.dump(store, f)
             print("Sleeping for 10 seconds and trying again")
             time.sleep(10)
             store.add_texts([i.page_content], metadatas=[i.metadata])
         c1 += 1
 
-    store.save_local("outputs")
+
+    faiss.write_index(store.index, "docs.index")
+    store.index = None
+    with open("faiss_store.pkl", "wb") as f:
+        pickle.dump(store, f)
 
 def get_user_permission(docs):
 # Function to ask user permission to call the OpenAI api and spend their OpenAI funds.