You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I used wandb to log the training process on an offline server A, and then I copied the logs to server B, which has internet access, to upload them using wandb sync. However, I encountered the following error, and the link is also a 404:
wandb sync .\offline-run-20250306_134443-8sfvkizj
Find logs at: C:\Users\lenovo\Downloads\offline-run-20250306_134443-8sfvkizj\mnt\dolphinfs\hdd_pool\docker\user\hadoop-aipnlp\INS\ruanjunhao04\verl-main\wandb\wandb\debug-cli.lenovo.log
Syncing: https://wandb.ai/a954369920/reinforce%2B%2Bacr-agi/runs/8sfvkizj ...
wandb: ERROR Error uploading "/home/hadoop-aipnlp/.local/share/wandb/artifacts/staging/tmpltuuv4qs": FileNotFoundError, [Errno 2] No such file or directory: '/home/hadoop-aipnlp/.local/share/wandb/artifacts/staging/tmpltuuv4qs'
wandb: ERROR Uploading artifact file failed. Artifact won't be committed.
wandb: ERROR Error uploading "/home/hadoop-aipnlp/.local/share/wandb/artifacts/staging/tmpulmc3fwf": FileNotFoundError, [Errno 2] No such file or directory: '/home/hadoop-aipnlp/.local/share/wandb/artifacts/staging/tmpulmc3fwf'
wandb: ERROR Uploading artifact file failed. Artifact won't be committed.
done.
The text was updated successfully, but these errors were encountered:
Hi @rangehow, the artifacts being logged and uploaded are store in a separate folder from the folder storing the wandb runs' data, which is the staging folder with the artifact file mentioned in the error
wandb: ERROR Error uploading "/home/hadoop-aipnlp/.local/share/wandb/artifacts/staging/tmpltuuv4qs": FileNotFoundError, [Errno 2] No such file or directory: '/home/hadoop-aipnlp/.local/share/wandb/artifacts/staging/tmpltuuv4qs'
So you won't be able to sync move the run and sync the artifact from a different device. If you have access to the original artifact files, I'd recommend to copy them across the device with access to wandb.ai, resume the run after syncing and log them that way.
I used
wandb
to log the training process on an offline server A, and then I copied the logs to server B, which has internet access, to upload them usingwandb sync
. However, I encountered the following error, and the link is also a 404:The text was updated successfully, but these errors were encountered: