stubbb
stubbb
Is this the one you are looking for: https://drive.google.com/drive/folders/1w9EjMkrjxOmMw3Rf6fXXkiv_ge7M99jR
I would not ever attempt to use this on anything below 2 GB of RAM.
Well, I reach 8GB of RAM to get that 10 fps at the accuracy I need while smoking 30 Watts at the most power efficient AI board on the market....
Is this the issue: **HW=368x432.onnx --w 368 --h 432** in your CLI use? As for your last example, its good to see that 2 images are produced correctly. The seg...
@salvador-blanco Can you tell me what operating system you use with the RTX 2080? I dont run these on RTX 2080. I think I know what causes your seg fault...
@ganler I can reproduce the segfaults when I use a large data set batched on a larger model. Try to reproduce with OpenPoseCOCO at something like 650x368 with 100+ images...
@paleomoon The lopps being 368x432 means actually 432x368. Flip your dimensions and should be ok.
Can you share how you invoke the model? I managed to correct this error every time I got it and I used this model.
Your network is actually broken: Network Input Shape: (-1, 3, -1, -1)
That might be the thing, I run it natively no issue. You can try to download the model, go into the container and swap it out.