converting yolov4-tiny.weights to tensorrt.engine

[revanb88]

Env

• GPU, e.g. V100, RTX2080, TX2, Xavier NX, Nano, etc.
• OS, e.g. Ubuntu16.04, Win10, etc.
• Cuda version
• TensorRT version

About this repo

• which branch/tag/commit are you using?
• which model? yolov5, retinaface?

Your problem

• what is your command? e.g. sudo ./yolov5 -s
• what's your output?
• what output do you expect?

[revanb88]

Hi, I am facing issue while converting yolov4-tiny.weights to tensorrt.engine following link https://github.com/wang-xinyu/tensorrtx/tree/master/yolov3-tiny GPU - V100 cuda - 11.4 ubuntu - 18.04

Traceback (most recent call last): File "gen_wts.py", line 12, in load_darknet_weights(model, weights) File "/home/brevanasiddappa/tensorrt-tiny/yolov3/models.py", line 392, in load_darknetweights conv.weight.data.copy(torch.from_numpy(weights[ptr:ptr + nw]).view_as(conv.weight)) RuntimeError: shape '[256, 384, 3, 3]' is invalid for input of size 695826

[wang-xinyu]

yolov4-tiny is not supported, you need to adapt it by yourself.

[revanb88]

thank you

On Tue, Oct 19, 2021 at 8:37 AM Wang Xinyu @.***> wrote:

yolov4-tiny is not supported, you need to adapt it by yourself.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/wang-xinyu/tensorrtx/issues/765#issuecomment-946327887, or unsubscribe https://github.com/notifications/unsubscribe-auth/AWDAXTMDLPRWSQV5VVHXWJLUHTOHFANCNFSM5GHBMDMA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

-- Thanks and Regards Dr. Revanasiddappa M Biradar Data Scientist Bengaluru

[MHGL]

@revanb88

# https://github.com/ultralytics/yolov3.git  models.py line 351-354
# set cutoff = 15
# will get yolov4-tiny.wts

but while convert wts to engine, i got this error:

[11/08/2021-17:11:33] [E] [TRT] Parameter check failed at: ../builder/Network.cpp::addScale::482, condition: shift.count > 0 ? (shift.values != nullptr) : (shift.values == nullptr)

[mingwz]

Hello, I have encountered a similar problem, and when I converted to yolov4, I got the following error: jeston nano cuda - 10.2 ubuntu - 18.04

Traceback (most recent call last): File "gen_wts.py", line 12, in load_darknet_weights(model, weights) File "/home/zmw/yolov3/models.py", line 392, in load_darknetweights conv.weight.data.copy(torch.from_numpy(weights[ptr:ptr + nw]).view_as(conv.weight)) RuntimeError: shape '[1024, 512, 3, 3]' is invalid for input of size 4559937 How should i solve it?

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.