Chris Van Pelt (CVP)
Chris Van Pelt (CVP)
Hey @chuber1986 would you be up for hopping on a quick screenshare today after 12pm PDT? I can help you do some deeper digging to see what the root cause...
Hey @chuber1986 apologies for the back and forth. I can do tomorrow at 11am PDT if that works for you? Zoom is preferred, but either work.
Glad we got it sorted. For anyone else coming to this thread, we're still digging into the root cause, but you can get yourself out of it with the following...
Whatever you choose, this is your account... This shouldn't be necessary however. If you're not accessing the local server at http://localhost:8080 you need to start the server with the host...
This likely means your database is corrupt. If you don't have any important data logged to the server, you can run: docker stop wandb-local docker volume rm wandb Then you...
You need to specify the HOST environment variable if you're not accessing the UI from 127.0.0.1 Can you share a screenshot or error you're seeing?
Sure! You can either override the entity by passing the -e flag to wandb sync or create a team with the same name as the old runs used and add...
If you change the file back to use the original entity name the checksum should succeed.
Hey @Jeffrey28 we just released a new version of the server that should address this problem. Can you try upgrading again?
@Jeffrey28 I've verified 0.22.0 works fine when upgrading from 0.20.0. There was a regression in 0.21.0. Can you try running: ``` docker pull wandb/local:latest docker stop wandb-local wandb server start...