Adding values manually to run

Hey!

I have some metrics which I cannot log during a run, but would like to attach to afterwards to also plot it. Is there a ways through the website or the CLI to manually attach a new value or metric?

Greetings,
Patrick

Hi @hallerpatrick,

Apologies about the delay here. You can not add metrics to a run’s history, just the final value of a metric. As such, I don’t believe metrics can be plotted this way.

Please let me know how your metrics are logged, and we can try to find a solution for how this can be done.

Thanks,
Ramit

​Hi @hallerpatrick,

We wanted to follow up with you regarding your support request as we have not heard back from you. Please let us know if we can be of further assistance or if your issue has been resolved.

Best,

Weights & Biases

Hi,

You can do something like this using the API.

import wandb
api = wandb.Api()

run = api.run("ayut/petfinder-train/ejcxhnb3")
run.summary["LB Score"] = 18.70218
run.summary.update()

Here ayut/petfinder-train/ejcxhnb3 is the run path