Use Mixtral

requirements.txt

@@ -2,3 +2,4 @@ langchain
 langchain-community
 langchain-huggingface
 chromadb
+InstructorEmbedding

worker.py

@@ -1,4 +1,3 @@
-import os
 import torch
 from langchain.chains import RetrievalQA
 from langchain_community.embeddings import HuggingFaceInstructEmbeddings
@@ -38,8 +37,9 @@ def init_llm():
 
     # repo name for the model
     # model_id = "tiiuae/falcon-7b-instruct"
-    model_id = "microsoft/Phi-3.5-mini-instruct"
+    # model_id = "microsoft/Phi-3.5-mini-instruct"
     # model_id = "meta-llama/Llama-3.2-1B-Instruct"
+    model_id = "mistralai/Mixtral-8x7B-Instruct-v0.1"
     
     # load the model into the HuggingFaceHub
     llm_hub = HuggingFaceEndpoint(repo_id=model_id, temperature=0.1, max_new_tokens=600, model_kwargs={"max_length":600})