[TF-TRT] DefaultLogger INVALID_ARGUMENT: Cannot set empty memory #52287
Labels
comp:gpu:tensorrt
Issues specific to TensorRT
stat:awaiting tensorflower
Status - Awaiting response from tensorflower
type:bug
Bug
System information
Describe the current behavior
Hi, while using TF-TRT, always I faced on an error
2021-10-04 00:54:11.478005: E tensorflow/compiler/tf2tensorrt/utils/trt_logger.cc:42] DefaultLogger INVALID_ARGUMENT: Cannot set empty memory.
when I calledconverter.build(input_fn=my_input_fn)
method. (either the error raised at runtime when not call build() method for pre-build engine.)(image : trials, parameter tuning
minimum_segment_size
and so on.)But any of them couldn't remove that error message.
My entire source code wasn't different with any other example source codes.
I found that optimized model does not improving model inference speed, so I guess the error is not just a warning. Maybe It's a bug.
Thanks!
Describe the expected behavior
Contributing
Standalone code to reproduce the issue
Provide a reproducible test case that is the bare minimum necessary to generate
the problem. If possible, please share a link to Colab/Jupyter/any notebook.
COLAB's TF-TRT is unstable.
Other info / logs Include any logs or source code that would be helpful to
diagnose the problem. If including tracebacks, please include the full
traceback. Large logs and files should be attached.
model signature
Entire program log description below
The text was updated successfully, but these errors were encountered: