diff --git a/examples/train_full/llama3_full_sft_ds3.yaml b/examples/train_full/llama3_full_sft_ds3.yaml
index 40b62f24..40afd2ee 100644
--- a/examples/train_full/llama3_full_sft_ds3.yaml
+++ b/examples/train_full/llama3_full_sft_ds3.yaml
@@ -5,9 +5,6 @@ model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
 stage: sft
 do_train: true
 finetuning_type: full
-
-### ddp
-ddp_timeout: 180000000
 deepspeed: examples/deepspeed/ds_z3_config.json
 
 ### dataset
@@ -33,6 +30,7 @@ num_train_epochs: 3.0
 lr_scheduler_type: cosine
 warmup_ratio: 0.1
 fp16: true
+ddp_timeout: 180000000
 
 ### eval
 val_size: 0.1