Hi, is there any tool/method I can check the parameters for tensorrt engine? like check the precision set during engine creation. Thanks.
NVES
April 27, 2022, 4:37am
2
Hi ,
These support matrices provide a look into the supported platforms, features, and hardware capabilities of the NVIDIA TensorRT 8.4.3 APIs, parsers, and layers.
You can refer below link for all the supported operators list.
<!--- SPDX-License-Identifier: Apache-2.0 -->
# Supported ONNX Operators
TensorRT 8.4 supports operators up to Opset 17. Latest information of ONNX operators can be found [here](https://github.com/onnx/onnx/blob/master/docs/Operators.md)
TensorRT supports the following ONNX data types: DOUBLE, FLOAT32, FLOAT16, INT8, and BOOL
> Note: There is limited support for INT32, INT64, and DOUBLE types. TensorRT will attempt to cast down INT64 to INT32 and DOUBLE down to FLOAT, clamping values to `+-INT_MAX` or `+-FLT_MAX` if necessary.
See below for the support matrix of ONNX operators in ONNX-TensorRT.
## Operator Support Matrix
| Operator | Supported | Supported Types | Restrictions |
|---------------------------|------------|-----------------|------------------------------------------------------------------------------------------------------------------------|
| Abs | Y | FP32, FP16, INT32 |
| Acos | Y | FP32, FP16 |
| Acosh | Y | FP32, FP16 |
| Add | Y | FP32, FP16, INT32 |
show original
Thanks!
Thanks for the prompt reply. I’m sorry that I might not clarify my question in a better way. My question is what if I have an engine where I forgot about the precision I have set during the engine creation (or any other parameters), is it possible for me to check back the precision that was set during the engine creation?
Hi,
By default, TensorRT chooses FP32 as a precision if you don’t mention it externally.trtexec verbose logs as well we can verify.
Thank you.