diff --git a/3_distributed_training/reinforcement-learning/grpo/trl/torchrun/fsdp/scripts/requirements.txt b/3_distributed_training/reinforcement-learning/grpo/trl/torchrun/fsdp/scripts/requirements.txt
index 1eafa43..b1ecb8d 100644
--- a/3_distributed_training/reinforcement-learning/grpo/trl/torchrun/fsdp/scripts/requirements.txt
+++ b/3_distributed_training/reinforcement-learning/grpo/trl/torchrun/fsdp/scripts/requirements.txt
@@ -8,7 +8,7 @@ huggingface_hub[hf_transfer]
 mlflow
 rouge
 safetensors>=0.5.2
-sagemaker==2.244.0
+sagemaker==2.256.0
 sagemaker-mlflow==0.1.0
 sentencepiece==0.2.0
 scikit-learn==1.6.1