Spaces:
Runtime error
Runtime error
ImranzamanML
commited on
Commit
•
be1b078
1
Parent(s):
e40592d
Update ai_assistant.py
Browse files- ai_assistant.py +27 -24
ai_assistant.py
CHANGED
@@ -1,3 +1,4 @@
|
|
|
|
1 |
from langchain.document_loaders.csv_loader import CSVLoader
|
2 |
from langchain.embeddings.openai import OpenAIEmbeddings
|
3 |
from langchain.embeddings import CacheBackedEmbeddings
|
@@ -5,49 +6,51 @@ from langchain_community.vectorstores import FAISS
|
|
5 |
from langchain.storage import LocalFileStore
|
6 |
from langchain.chains import RetrievalQA
|
7 |
from langchain_openai import ChatOpenAI
|
8 |
-
import os
|
9 |
|
10 |
def create_index():
|
11 |
-
#
|
12 |
-
|
13 |
-
|
14 |
-
loader = CSVLoader(file_path = df_path)
|
15 |
-
data = loader.load()
|
16 |
|
17 |
-
#
|
18 |
embeddings_model = OpenAIEmbeddings()
|
19 |
|
20 |
-
#
|
21 |
store = LocalFileStore("./cache")
|
22 |
-
|
23 |
embeddings_model, store, namespace=embeddings_model.model
|
24 |
)
|
|
|
|
|
25 |
vector_store = FAISS.from_documents(data, embeddings_model)
|
26 |
|
27 |
return vector_store.as_retriever()
|
28 |
|
29 |
-
def
|
30 |
-
|
31 |
os.environ["OPENAI_API_KEY"] = openai_key
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
|
|
|
|
|
|
37 |
|
38 |
-
|
39 |
-
|
|
|
|
|
40 |
# Create the QA chain
|
41 |
handler = StdOutCallbackHandler()
|
42 |
-
|
43 |
qa_with_sources_chain = RetrievalQA.from_chain_type(
|
44 |
-
llm=
|
45 |
retriever=retriever,
|
46 |
callbacks=[handler],
|
47 |
return_source_documents=True
|
48 |
)
|
49 |
|
50 |
-
# Ask a question
|
51 |
-
res = qa_with_sources_chain({"query":query})
|
52 |
-
return
|
53 |
-
|
|
|
1 |
+
import os
|
2 |
from langchain.document_loaders.csv_loader import CSVLoader
|
3 |
from langchain.embeddings.openai import OpenAIEmbeddings
|
4 |
from langchain.embeddings import CacheBackedEmbeddings
|
|
|
6 |
from langchain.storage import LocalFileStore
|
7 |
from langchain.chains import RetrievalQA
|
8 |
from langchain_openai import ChatOpenAI
|
|
|
9 |
|
10 |
def create_index():
|
11 |
+
# Load the data from CSV file
|
12 |
+
data_loader = CSVLoader(file_path="train.csv")
|
13 |
+
data = data_loader.load()
|
|
|
|
|
14 |
|
15 |
+
# Create the embeddings model
|
16 |
embeddings_model = OpenAIEmbeddings()
|
17 |
|
18 |
+
# Create the cache backed embeddings in vector store
|
19 |
store = LocalFileStore("./cache")
|
20 |
+
cached_embedder = CacheBackedEmbeddings.from_bytes_store(
|
21 |
embeddings_model, store, namespace=embeddings_model.model
|
22 |
)
|
23 |
+
|
24 |
+
# Create FAISS vector store from documents
|
25 |
vector_store = FAISS.from_documents(data, embeddings_model)
|
26 |
|
27 |
return vector_store.as_retriever()
|
28 |
|
29 |
+
def setup_openai(openai_key):
|
30 |
+
# Set the API key for OpenAI
|
31 |
os.environ["OPENAI_API_KEY"] = openai_key
|
32 |
+
|
33 |
+
# Create index retriever
|
34 |
+
retriever = create_index()
|
35 |
+
|
36 |
+
# Initialize ChatOpenAI model
|
37 |
+
chat_openai_model = ChatOpenAI(model="gpt-4")
|
38 |
+
|
39 |
+
return retriever, chat_openai_model
|
40 |
|
41 |
+
def ai_doctor_chat(openai_key, query):
|
42 |
+
# Setup OpenAI environment
|
43 |
+
retriever, chat_model = setup_openai(openai_key)
|
44 |
+
|
45 |
# Create the QA chain
|
46 |
handler = StdOutCallbackHandler()
|
|
|
47 |
qa_with_sources_chain = RetrievalQA.from_chain_type(
|
48 |
+
llm=chat_model,
|
49 |
retriever=retriever,
|
50 |
callbacks=[handler],
|
51 |
return_source_documents=True
|
52 |
)
|
53 |
|
54 |
+
# Ask a question/query
|
55 |
+
res = qa_with_sources_chain({"query": query})
|
56 |
+
return res['result']
|
|