Luca Lazzaroni
Luca Lazzaroni
I think the problem lies in the Carla server build, I now build in Shipping mode using the dev branch and the memory leak seems milder even if it still...
Same issue here.
Managed to solve by adding `export VK_ICD_FILENAMES="/usr/share/vulkan/icd.d/nvidia_icd.json"` to my `.bashrc`
Thanks for the infos, I'll give it a try in the next few days. Meanwhile I opened the notebook locally with Jupyter but when it executes the line importing tensorflow...
I followed the steps you suggested but the error persists, So i tried to import tensorflow outside jupyter but it returns me `Illegal instruction (core dumped)`. At this point I...
Hi @AdamGleave, sorry for my late response, I tried to compile tensorflow from source but no luck. I think I'll try again with colab and see if there's any possibility...
Does the patch still work on the newest Carla version?
@ambitious-octopus thank you for the quick reply. Actually, I am using the same exact code but I am not able to generate any int8 model. On my Windows machine the...
@ambitious-octopus thank you, it was indeed a data issue because with `coco8.yaml` it works. However, I don't know how can it be possible that it's a data problem because I...
@ambitious-octopus I managed to find the issue. It seems that the process crashes when the number of images in the representative dataset is too big. In my case I had...