Hi,
Since TensorRT picks different algorithms based on the hardware/software resources.
So it’s possible that it picks some algorithm that contains larger precompute data.
If the inference output runs okay, it should be fine.
Thanks.
AastaLLL
3
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| The same onnx file produces engine files with different size under same configurations | 4 | 695 | May 12, 2022 | |
| TX2 NX ONNX Convert TensorRT Engine | 2 | 622 | October 18, 2021 | |
| Trt file from onnx is too large | 1 | 911 | March 10, 2021 | |
| Serialized trt engine file is much larger than the original model file? | 2 | 874 | October 12, 2021 | |
| TRT engine file bigger than original Darknet YOLO V3 model | 3 | 566 | January 30, 2023 | |
| Issue regrading size and type of the model during conversion from ONNX to TRT | 2 | 436 | October 18, 2021 | |
| Model onnx trt engine generation process report different results compared between PC and jetson XAVIER NX | 19 | 1055 | September 28, 2022 | |
| Why is the model engine file so much large with TensorRT 7.1 compared to 5.1 or 6.0? | 5 | 523 | July 8, 2020 | |
| File size vs. batch size for TensorRT serialized engine files | 1 | 569 | September 30, 2020 | |
| Model file size on jetson nano with 6.0-full-dims is larger than that on desktop PC with 6.0 | 4 | 768 | October 18, 2021 |