Update app.py

app.py

@@ -64,7 +64,7 @@ if compute_dtype == torch.float16 and use_4bit:
 
 model = AutoModelForCausalLM.from_pretrained(
     model_name, quantization_config=bnb_config)
-stop_list = [" \n\nAnswer:",  " \n\n"]
+stop_list = [" \n\nAnswer:", " \n ",  " \n\n"]
 stop_token_ids = [tokenizer(x,  return_tensors='pt', add_special_tokens=False)['input_ids'] for x in stop_list]
 stop_token_ids = [torch.LongTensor(x).to("cuda") for x in stop_token_ids]
     
@@ -97,8 +97,8 @@ db3 = Chroma(persist_directory="chroma/", embedding_function=instructor_embeddin
 
 
 
-retriever = db3.as_retriever(search_type="similarity_score_threshold",
-                                 search_kwargs={"score_threshold": .8,
+retriever = db3.as_retriever(search_type="mmr",
+                                 search_kwargs={
                                                 "k": 20})
 
 #retriever = db3.as_retriever(search_kwargs={"k":15})