diff --git a/nemo/export/tensorrt_llm.py b/nemo/export/tensorrt_llm.py index b38eb592529c..d987eca2a7a8 100644 --- a/nemo/export/tensorrt_llm.py +++ b/nemo/export/tensorrt_llm.py @@ -245,6 +245,7 @@ def build( max_input_len: int = 256, max_output_len: int = 256, max_batch_size: int = 8, + gpus_per_node: int = 8, use_refit: bool = False, reshard_model: bool = False, ):