Axel Mendoza

Hi @timmeinhardt, I'm still stuck and can't fix the problem. Any idea about xhat to do ?

The overall loss before crash is: ```bash 7.164515495300293 4.680170059204102 4.274537086486816 5.886708736419678 4.5304131507873535 3.8330774307250977 7.964834690093994 3.0583243370056152 4.674091339111328 3.301724672317505 3.608858346939087 2.8133819103240967 2.8961634635925293 2.8605575561523438 3.1516401767730713 3.2725605964660645 2.9872331619262695 1.8360798358917236 2.4671669006347656 ``` I logged this...

Unfortunately, because the training crashes very early, the losses do not appear in visdom.

Nice, that worked.  But yeah, it looks like the losses are not the issue. I have no idea what to do ;(

Except that there is a lot of region proposal, it seems fine. The groundtruths are well placed. I just tried the deactivate all augmentations an I had the same results....

Hi @nmanovic, Thank you for the quick response. It will be very useful for us to be able to access some statistics using the REST API. I'm currently working on...

Nice fix. It worked for me with the same configuration and on GCP too. Thanks for you're help !

Hi @bsekachev, Here it is: ```bash 2022-11-08 13:54:04,112 DEBG 'runserver' stderr output: [Tue Nov 08 13:54:04.112792 2022] [wsgi:error] [pid 68:tid 140604987524864] [remote 192.168.224.4:50008] [2022-11-08 13:54:04,112] INFO cvat.server.task_285: get repository request...

@bsekachev ```bash 2022-11-08 13:54:24,611 DEBG 'rqworker_default_1' stderr output: DEBUG:rq.worker:Handling failed execution of job /api/task/285/annotations/COCO Keypoints 1.0 2022-11-08 13:54:24,612 DEBG 'rqworker_default_1' stderr output: ERROR:rq.worker:Traceback (most recent call last): File "/opt/venv/lib/python3.8/site-packages/datumaro/plugins/coco_format/converter.py", line...

@zhiltsov-max I have no issue when I export with the format 'CVAT for videos 1.1' with the 'Save Images Option'. Whereas the issue happens with the 'COCO Keypoints 1.1' format.