Yesterday, wandb initialization was working for me both from my local desktop and on a remote EC2 instance. Today, I get the following error:
**wandb** : Network error (ReadTimeout), entering retry loop. See wandb/debug-internal.log for full traceback.
I haven’t changed anything about the settings, so this makes me think this is a wandb service issue. Is there a page on the website that tells the status of wandb services?
We are currently experiencing some issues with degraded performance, which is likely the cause of the error. We are aware of this problem, and are trying to fix it as soon as possible. You can check here to check the status of our service.
As a workaround for now, you can run wandb fully offline, and sync your files later using wandb sync. Since you will be training offline, no network requests will be made.
@ramit_goolry I encountered the same issue, and at the time the status page hadn’t been updated (despite getting the aforementioned timeout message for about an hour).
It would be useful to have it either updated faster, or have status updates posted on Twitter (or equivalent). I spent about 2 hours trying to figure out the issue on our side before realising it was an issue on the W&B side after seeing a random comment on Twitter.
I apologize for the delay for the update. I will let the engineering team know that the status page needs to be updated more frequently in case of an outage.