Open kristofkammerhofer2 opened 4 weeks ago
👋 Hello @kristofkammerhofer2, thank you for raising this issue about Ultralytics HUB 🚀! This is an automated response to guide you while an Ultralytics engineer looks into this shortly.
Here's how you can explore more about HUB and triage common issues:
Since you've encountered a 🐛 issue, could you help us by providing a minimum reproducible example (MRE)? This includes screenshots and a detailed step-by-step process to reproduce the issue. You can learn more about how to provide an MRE here.
We appreciate your patience and will respond as soon as we can! 😊
Thank you, and stay tuned for more assistance from our team!
@kristofkammerhofer2 thank you for the bug report! We have a known issue with GPU training and are looking into this.
@glenn-jocher - ok, if it's a known issue, how much longer will fixing it take? Why do you accept topping up the account then? And why are you not flagging it when users are trying to do it, wasting hours of my time?
@glenn-jocher - when is it going to be fixed? If you guys are too busy with other stuff, just refund me.
@kristofkammerhofer2 Do you still have this issue? It should be fixed now. If you still have this issue, try re-creating the model (using the Train Model button).
Search before asking
HUB Component
Training
Bug
After successfully setting up a cloud instance to start training the website tells me that it's initiating for about 10 minutes and then reverts back to the original state where I need to start the training again.
So the training is not happening. I've tried different GPUs so far but nothing works.
What can I do?
Environment
No response
Minimal Reproducible Example
No response
Additional
No response