diff --git a/nemo/export/tensorrt_llm.py b/nemo/export/tensorrt_llm.py index b2685f4dc50b..86023fa079e6 100644 --- a/nemo/export/tensorrt_llm.py +++ b/nemo/export/tensorrt_llm.py @@ -171,7 +171,7 @@ def export( paged_kv_cache: bool = True, remove_input_padding: bool = True, paged_context_fmha: bool = False, - dtype: Optional[str] = "bfloat16", + dtype: Optional[str] = None, load_model: bool = True, use_lora_plugin: str = None, lora_target_modules: List[str] = None,