导入torch2trt时发生属性错误

时间:2020-06-06 16:06:25

标签: python-3.6 tensorrt nvidia-jetson-nano

我在jetson nano上使用torch2trt。我已经使用官方文档安装了带有插件的torch2trt。我正在使用最新的jetpack和版本(7.1.0.16)的Tensorrt。我收到以下错误。

import torch2trt
Traceback (most recent call last):
File "", line 1, in 
File "/home/dev/.virtualenvs/pytorch/lib/python3.6/site-packages/torch2trt/init.py", line 1, in 
from .torch2trt import *
File "/home/dev/.virtualenvs/pytorch/lib/python3.6/site-packages/torch2trt/torch2trt.py", line 5, in 
from .calibration import TensorBatchDataset, DatasetCalibrator, DEFAULT_CALIBRATION_ALGORITHM
File "/home/dev/.virtualenvs/pytorch/lib/python3.6/site-packages/torch2trt/calibration.py", line 5, in
if trt.version >= '5.1':
AttributeError: module 'tensorrt' has no attribute 'version'

0 个答案:

没有答案