diff --git a/finetune/finetune_lora.sh b/finetune/finetune_lora.sh
index df3140a4..c57409fa 100644
--- a/finetune/finetune_lora.sh
+++ b/finetune/finetune_lora.sh
@@ -53,7 +53,7 @@ torchrun $DISTRIBUTED_ARGS finetune.py  \
     --per_device_train_batch_size 1 \
     --per_device_eval_batch_size 1 \
     --gradient_accumulation_steps 1 \
-    --evaluation_strategy "steps" \
+    --eval_strategy "steps" \
     --save_strategy "steps" \
     --save_steps 1000 \
     --save_total_limit 10 \