requirement.txt

app.py

@@ -3,19 +3,21 @@ import os
 import pinecone
 import time
 # from torch import cuda
-from langchain.embeddings.huggingface import HuggingFaceEmbeddings
+# from langchain.embeddings.huggingface import HuggingFaceEmbeddings
 # import PyPDF2
 # import re
 from langchain.vectorstores import Pinecone
+from sentence_transformers import SentenceTransformer
 
-embed_model_id = 'sentence-transformers/all-MiniLM-L6-v2'
+
+embed_model = SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')
 # device = f'cuda:{cuda.current_device()}' if cuda.is_available() else 'cpu'
 
-embed_model = HuggingFaceEmbeddings(
-    model_name=embed_model_id,
-    # model_kwargs={'device': device},
-    # encode_kwargs={'device': device, 'batch_size': 32}
-)
+# embed_model = HuggingFaceEmbeddings(
+#     model_name=embed_model_id,
+#     # model_kwargs={'device': device},
+#     # encode_kwargs={'device': device, 'batch_size': 32}
+# )
 
 # get API key from app.pinecone.io and environment from console
 pinecone.init(

requirements.txt

@@ -6,4 +6,5 @@ langchain==0.0.240
 datasets==2.14.0
 accelerate==0.21.0
 einops==0.6.1
-faiss-cpu
+faiss-cpu
+sentence_transformers