Update README.md

README.md

@@ -35,7 +35,7 @@ tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
 model = MixtralForCausalLM.from_pretrained(model_name_or_path, load_in_8bit=True)
 
 # set num_experts_per_tok 1 or 2 ?
-model.config.num_experts_per_tok = 1
+model.config.num_experts_per_tok = 2
 
 # message
 messages = [