Detectron2 needs Cuda max 11.3 Version

holge · March 10, 2022, 5:02pm

Hi guys

we just installed a new RTX3070. Distribution is Ubuntu 18.04 LTS.
The driver installation was successful with the latest driver 510.47.03.
Unfortunetly the cuda version is 11.6. It is not compatible with detectron 2 installation.
Pytorch 1.10 is requested and max cuda version 11.3.
I am not able to downgrade the version to 11.3.
Installation of an older driver did not work. Only cuda installation did not work either.
Please give some installation hints how to downgrade the cuda version.
BR Holger

NVES · March 10, 2022, 5:38pm

Hi ,
We recommend you to check the supported features from the below link.

You can refer below link for all the supported operators list.
For unsupported operators, you need to create a custom plugin to support the operation

github.com

onnx/onnx-tensorrt/blob/main/docs/operators.md

<!--- 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 |

This file has been truncated. show original

Thanks!

Topic		Replies	Views
Is there any way to install a higher version than 8.0 of TensorRT through tar without changing the cuda11.3 version? TensorRT	1	313	June 9, 2023
Is there any way to upgrade tensorrt inside official docker container? TensorRT	1	632	February 23, 2023
TensorRT version for CUDA 12.0 TensorRT tensorrt , cuda	1	1750	March 14, 2023
The detected CUDA version (12.0) mismatches the version that was used to compile PyTorch (11.4). Please make sure to use the same CUDA versions TensorRT tensorrt , cuda , python	1	8542	August 7, 2023
nVidia release versions compatibility TensorRT	4	913	September 20, 2023
Suitable cuda toolkit and cudnn for windows laptop TensorRT cuda	1	496	July 2, 2023
TensorRT 8.2 + CUDA 11.4 -> which CuDNN version? TensorRT	6	3168	November 24, 2021
After removing Tensorrt, can't install anything NVIDIA TensorRT cudnn	0	372	December 20, 2023
Unable to use TensorRT 7.2.3 for Ubuntu 18.04 and CUDA 11.2 TensorRT	6	4547	May 6, 2021
TensorRT CUDNN CUDA compatibility issue TensorRT tensorrt , cuda , gpu	1	4392	April 25, 2022

Detectron2 needs Cuda max 11.3 Version

Related topics