David
David
Hi all, I have a first version for the the group_norm layer converter using the GN plugin in TensorRT. I add a group_norm.cpp file that wrap the correct pytorch signature....
I have patch the group_norm from PyTorch without the check of the batch size, but it seems that with the ngc21.12 there is no need to do that
The GroupNormPlugin provided by TensorRT doesn't support the fine-tuned weights and biases. I bypass this missing feature with the addScaleNd TensorRT layer that I plugged after the call of the...
Thanks! If this is the case it would be great to have a clear message that indicates if an ONNX attributes is not supported by the ONNX2TensorRT driver
Do you mean that you will fill a feature request to support antialias mode in resize?
Hello, any news on supporting Antialiasing in the resize operators? Else, would you have a plugin example to share?
Having any guideline on how adapting a BF16 trained model for a TensorRT (version < 9) would be super useful in the meantime
In addition, I would like to know if there is a way to access the cubin file for the multi-heads cross-attention plugin for head dimension == 32. Or even the...
Is it compatible with [Torch-TensorRT](https://pytorch.org/TensorRT/)?
@nvpohanh thanks for the link. When I compile a model with the trtexec executable 9.2 (+ all the libraries installed) I encounter this issue when I try to execute the...