Could not find any implementation for node {ForeignNode[/model/Unsqueeze + /model/Unsqueeze_1.../Cast]} failure of TensorRT 10.13.3 when running trtexec on GPU 3090

Description

Could not find any implementation for node {ForeignNode[/model/Unsqueeze + /model/Unsqueeze_1.../Cast]} failure of TensorRT 10.13.3 when running trtexec on GPU 3090

Environment

TensorRT Version: 10.13.3

NVIDIA GPU: RTX 3090

NVIDIA Driver Version: 580.82.07

CUDA Version: 13.0

CUDNN Version: 8.9.7.29

Operating System: Ubuntu 22.04

Python Version (if applicable): 3.10

Tensorflow Version (if applicable):

PyTorch Version (if applicable):

Baremetal or Container (if so, version):

Relevant Files

Model link: https://github.com/lchi021497/LLaVA

Steps To Reproduce

Commands or scripts:

• run trtexec --onnx=llava.onnx --saveEngine=llava.engine --builderOptimizationLevel=0 in the container from this repo with the ./docker/launch.sh command.

Have you tried the latest release?: yes

Attach the captured .json and .bin files from TensorRT's API Capture tool if you're on an x86_64 Unix system

Can this model run on other frameworks? For example run ONNX model with ONNXRuntime (polygraphy run <model.onnx> --onnxrt):

Start with convert.py and the docker/launch.sh environment, then reproduce the failure using trtexec --onnx=llava.onnx --saveEngine=llava.engine --builderOptimizationLevel=0. Inspect the reported GPU allocation failure and ForeignNode[/model/Unsqueeze + /model/Unsqueeze_1.../Cast] error. Done means determining whether the conversion fails from insufficient memory, unsupported graph nodes, or both, with a reproducible diagnosis.