diff --git a/src/python/py/models/builders/base.py b/src/python/py/models/builders/base.py index 83b06e4672..25a5f238d7 100644 --- a/src/python/py/models/builders/base.py +++ b/src/python/py/models/builders/base.py @@ -1968,6 +1968,7 @@ def make_rotary_embedding_multi_cache(self, **kwargs): sin_cache_small_name=sin_cache_small_name, small_cache_shape=cos_cache_large.shape, ) + self.ep_attrs["trt-rtx"]["enable_cuda_graph"] = "0" return # For other EPs (CUDA, CPU, WebGPU), create regular If node with multiple outputs