Description
I have two onnx models, one is the base model and the other is the small lmodel, which is a part of the base model and fully included in it. Both models have a combination of operators including gridsample and two reducemax. I found that when performing trtexec on these two models, the fusion method of this operator combination is different. The fusion method of the base model is to fuse gridsample+other operators together, and the two reducemax are fused together; The fusion method of the small model is to merge gridsample+the first reducemax+other operators together, and the second reducemax+other operators together. I want to know how to control this fusion way
Environment
TensorRT Version: v101010
GPU Type: Thor
Nvidia Driver Version:
CUDA Version:
CUDNN Version:
Operating System + Version:
Python Version (if applicable):
TensorFlow Version (if applicable):
PyTorch Version (if applicable):
Baremetal or Container (if container which image + tag):
Relevant Files
Please attach or include links to any models, data, files, or scripts necessary to reproduce your issue. (Github repo, Google Drive, Dropbox, etc.)
Steps To Reproduce
Please include:
- Exact steps/commands to build your repro
- Exact steps/commands to run your repro
- Full traceback of errors encountered