Switched LLM to mistralai/Mistral-7B-Instruct-v0.2.

app/server.py

@@ -14,7 +14,7 @@ app = FastAPI()
 # os.environ['TRANSFORMERS_CACHE'] = '/blabla/cache/'
 
 hf_llm = HuggingFaceHub(
-    repo_id="mistralai/Mixtral-8x7B-v0.1",
+    repo_id="mistralai/Mistral-7B-Instruct-v0.2",
     huggingfacehub_api_token=os.environ["HF_TOKEN"],
     task="text-generation",
     model_kwargs={"temperature":0.01, "max_new_tokens" : 250}