[CLI]: ChunkedEncodingError: Response ended prematurely about wandb

[2er0]

I was able to collect the results in a different way. We can close this issue from my side.
I'm not pulling a CSV via the API.

[ArtsiomWB]

Hi @2er0! Thank you for writing in!

What particular metadata do you usually need to pull from wandb API. Could you please provide a workspace where you are seeing a behavior of the table download button as cvs not downloading all runs?

[ArtsiomWB]

Hi there, I wanted to follow up on this request. Please let us know if we can be of further assistance or if your issue has been resolved.

[2er0]

The issue still exists.
I have one workspace where it works and one where the error above occurs:

• not working: timeflow/timeflow-mTADS-4-idun-real
• working: timeflow/timeflow-mTADS-2-idun

[ArtsiomWB]

Thank you for sending over the example workspaces, that is very helpful.

When you run your script above, does it crash right away, or after pulling some data from wandb?

[2er0]

The script pulls ~117 runs and then gets the error seen above.

[ArtsiomWB]

Hi @2er0, apologies it took a while to get back to you, this thread got completely burried. What would you like your CSV to look like while pulling it from the API?

Also does it always error out at run 117?