Is there a docker tar package that can be used directly?
You can download the docker image here.
Thanks for the amazing work @DerryHub!
Seems like the Docker image link is expired. Can you share an updated link? I will recommend Google Drive in case the links on this file sharing platform come with an expiry date.
P.S. already tried the readme link too, seems to be the same link.
Try this one https://pan.baidu.com/s/1dPR6kvgpUoKow51870KNug?pwd=6xkq
Can inference use trtexec directly?
create docker has meet some problem, Such:
docker: Error response from daemon: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: exec: "/bin/bash": stat /bin/bash: no such file or directory: unknown.
I upload it to docker hub and fix some minor bugs.
Just elegantly try the following script
$docker run -it --rm \
--gpus all \
-v ./:/workspace/BEVFormer_tensorrt/ \
-v /data/nuscenes/can_bus:/workspace/BEVFormer_tensorrt/data/can_bus \
-v /data/coco:/workspace/BEVFormer_tensorrt/data/coco \
-v /data/nuscenes:/workspace/BEVFormer_tensorrt/data/nuscenes \
--shm-size=16G \
--privileged \
--network=host \
--user root \
--env HTTPS_PROXY= \
hadonga/bev_trt:1.0 /bin/bash
I upload it to docker hub and fix some minor bugs. Just elegantly try the following script
$docker run -it --rm \ --gpus all \ -v ./:/workspace/BEVFormer_tensorrt/ \ -v /data/nuscenes/can_bus:/workspace/BEVFormer_tensorrt/data/can_bus \ -v /data/coco:/workspace/BEVFormer_tensorrt/data/coco \ -v /data/nuscenes:/workspace/BEVFormer_tensorrt/data/nuscenes \ --shm-size=16G \ --privileged \ --network=host \ --user root \ --env HTTPS_PROXY= \ hadonga/bev_trt:1.0 /bin/bash
Hello, can this image run under jetson orin nx?
Try this one https://pan.baidu.com/s/1dPR6kvgpUoKow51870KNug?pwd=6xkq
I encountered below question when I tried to run pth2onnx.py in this docker image "ImportError: cannot import name 'version' from 'torch.torch_version' (/usr/local/lib/python3.8/dist-packages/torch/torch_version.py)"
Do you have the docker image pushed on dockerhub? it is easy to docker pull from the docker hub