Training stuck at "Optimizing Hyperparameters" for 1-2 hours, then "Training Failed"

After I start training (free account, still 3 free tokens left) it does some stuff then gets to “Optimizing Hyperparameters…”, where it gets stuck for a hour or two, then this appears:

Are the roboflow servers down or what?

  • Project Type: Object detection
  • Operating System & Browser: Windows & Google Chrome
  • Project Universe Link or Workspace/Project ID: brawl-stars-character-circles

This is the error that appears afer training fails in a hour or two:


This is the devtools

Training the default hard-hat-sample-flu1v/1 seems to work fine, but making my own model with only pre proccesing to decrease resolution and 93 images (of which 68 for training) still fails