-
Notifications
You must be signed in to change notification settings - Fork 640
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[CLI]: wandb makes no progress uploading artifacts after hours of waiting #7266
Comments
Possibly related: after running into this issue with two wandb processes simultaneously, I'm seeing
|
Seems to be fixed after upgrading to the latest wandb, but it's pretty unfriendly to hang forever in recent versions of the CLI |
@JasonGross sorry that you experienced this issue and glad to hear it was resolved. Agreed if we didn't provide an appropriate indication it is not great. |
Describe the bug
I see
This is for https://wandb.ai/gbmi/MaxOf2-3000-epochs-adj-0,1,2,17-training-ratio-0.100-with-eos-nondeterministic/runs/ku05a7r7. The media does not display at all, despite uploading plotly.express plots on every epoch. It has been stuck at 0 MB for hours. I am not 100% sure how to reproduce, but you could try running
poetry run python -m gbmi.exp_max_of_n.train --max-of 2 --non-deterministic --train-for-epochs 3000 --validate-every-epochs 1 --force-adjacent-gap 0,1,2,17 --use-log1p --lr 0.001 --betas 0.9 0.98 --weight-decay 1.0 --optimizer AdamW --training-ratio 0.099609375 --log-matrix-interp --checkpoint-every-epochs 1 --batch-size 408 --log-every-n-steps 1 --use-end-of-sequence --use-kaiming-init
from https://github.com/JasonGross/guarantees-based-mechanistic-interpretability/Additional Files
No response
Environment
WandB version: 0.16.3
OS: Linux 0e5569e59be2 5.4.0-172-generic #190-Ubuntu SMP Fri Feb 2 23:24:22 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux
Python version: 3.10.13
Versions of relevant libraries: ?
Additional Context
No response
The text was updated successfully, but these errors were encountered: