Update app.py

app.py

@@ -21,11 +21,10 @@ from langchain.chains import LLMChain
 from langchain_core.runnables import RunnablePassthrough, RunnableParallel
 
 
-instructor_embeddings = HuggingFaceEmbeddings(model_name="intfloat/multilingual-e5-large",
-                                                      model_kwargs={"device": "cuda"})
 
+model_name= 'mistralai/Mistral-7B-v0.1'
 
-model_name='SherlockAssistant/Mistral-7B-Instruct-Ukrainian'
+#model_name='SherlockAssistant/Mistral-7B-Instruct-Ukrainian'
 
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 #tokenizer.pad_token = tokenizer.unk_token
@@ -94,6 +93,7 @@ text_generation_pipeline = pipeline(
 )
 mistral_llm = HuggingFacePipeline(pipeline=text_generation_pipeline)
 # # # load chroma from disk
+instructor_embeddings = HuggingFaceEmbeddings(model_name="intfloat/multilingual-e5-large")
 db3 = Chroma(persist_directory="/chroma/", embedding_function=instructor_embeddings)