-
Notifications
You must be signed in to change notification settings - Fork 2.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update yolox.cpp: fix INPUT_BLOB_NAME & OUTPUT_BLOB_NAME #1786
Conversation
LGTM. |
This change does not work with torch2trt converted models. |
Any extra information such as torch2trt version, error log etc. |
Hi @starlipernl I was not able to install trt2torch, in order to convert the model, I used : Best Regards, |
I used the latest release of both YOLOX and torch2trt with Tensorrt version 8.6.1. When trying to run the converted model with the cpp demo script I got the following error: This is because torch2trt defaults to input_0 and output_0 for the binding names. There may be a way to specific the input and output name bindings during torch2trt conversion within the trt.py file in this repo I just did not dig too much into it. |
I fixed it by just reverting the yolox.cpp script back to the original input and output names (input_0, output_0) |
I have already reverted this pr in #1792. |
Hi,
This merge request update the name of input & output blob name.
It allow to run tensorrt onnx models converted using
trtexec --onnx=yolox_s.onnx --saveEngine=yolox_s.trt
Best Regards,
Michel.