Visual recognition training failed

Question

I'm trying to train custom classifier, but still getting failed status.

{
    "classifier_id": "Castles_1969040174",
    "name": "Castles",
    "status": "failed"
}

I checked images and they are in right size (< 10 MB each) and both ZIPs are containing more than 10 images and are way smaller then 100 MB.

Is here any other condition I could check to get my classifier to be trained correctly?

Not sure how your request looks like. Can you add your request here? Also, I am assuming you are passing positive and negative examples for the proper training of your classifier. — Vidyasagar Machupalli
Thanks for comment, at the end, I end up with setting up all from beginning and now it look OK :-) — Sk1X1

chughts chughts · Accepted Answer · 2019-04-23T14:22:39

Run code to retrieve classifier details, check the API documentation for how -

for curl it will look like

curl -u "apikey:{apikey}" "https://gateway.watsonplatform.net/visual-recognition/api/v3/classifiers/Castles_1969040174?version=2018-03-19"

The response will contain an explanation telling you why the create classifier failed.