Error help: WARN: failed to process 158133/training/Loud sorl2.wav.3i23qleq.ingestion-f98b8fd84-mjnqg.json: payload.values.push is not a function

**Question/Issue: Hi! I am trying to make an audio detection model for an art project in school, and I have loaded in audio data as usual. I am getting this error when generating features, and my “sorl” class of audio data is not present after it is done. Here is the error message, any ideas what is wrong here?

WARN: failed to process 158133/training/Loud sorl2.wav.3i23qleq.ingestion-f98b8fd84-mjnqg.json: payload.values.push is not a function

This seems to happen for all my “sorl” class file (probably), but not the other two classes (unknown and noise). The error log gets flooded and shuts down. After finishing the generate features process, only “unknown” and “noise” is visible in the data visualization.

Cheers,
J

I also tried to export the data and the reuploading it again, but the export link does not work, amazon seems to be having issues?

Will prob try restarting from scratch again and resplit all my longer data, but I was hoping I wouldnt have to. Hope someone else can help me figure out the problem! Thanks.
/J

Hello @Instant_exit,

Could you provide your project ID so I can have a look please?

Best,

Louis

Im sorry, I had to reupload the data, and then it seems to be working (however I am still getting errors that the hash is already existing so it skips a few hundred files for some reason). But then I accidently relabeled my data just now (the label filter seems to be not working so all files got selected instead of only those of the filtered label?) and had to start over. So unfortunately there is no project with this error you can have a look at atm.

Sorry…
/J

Actually.

I think there is some sort of bug for the filtering system. I select my samples with filters to change the label for about half of them by using the filename as my filter, but all of the samples gets new labels, even though the system says not all of them are selected. Had to restart my model several times because of this and reupload all data. And then try to not have to relabel anything that is alot of files to avoid this issue.

/J