Skip to Content
1

Training ML models on SCP-CF

Feb 22 at 03:41 PM

63

avatar image
Former Member

Hi All,

I'm in the middle of developing an app for image recognition, using tensorflow at its core to train the model.


While parsing this project, I stumbled upon a problem and thus the question related to that is specified below.


Question: The BYOM documentation explains how to deploy an already trained model to the CF environment. But is it possible to create gpu-enabled instances in SCP-CF environment to train models?

10 |10000 characters needed characters left characters exceeded
* Please Login or Register to Answer, Follow or Comment.

1 Answer

Best Answer
Moya Watson
Feb 23 at 03:24 PM
1

Hi Sagar!

Thanks for writing up this question. We’ve had a discussion with the experts internally and I learned that while not available at present, train-your-own-model is in fact on the roadmap, tentatively set for later this year —- so stay tuned!

Can you reach this roadmap link, by the way? https://www.sap.com/registration/protected/default-overlay.html?gated_asset_path=%2Fdocuments%2F2017%2F06%2F223b68c7-c37c-0010-82c7-eda71af511fa.gate.html#pdf-asset=223b68c7-c37c-0010-82c7-eda71af511fa&page=3

Does this address your question?

Best,

-m

Show 3 Share
10 |10000 characters needed characters left characters exceeded
Former Member

Hi Moya,


Thanks for going deep on this and helping out.
Yes, this does answer my question. Looking forward to this feature in the future release. Thanks.

- Sagar J.

1
Former Member

To answer your other question - yes, I'm able to access the roadmap link.

1

Super! Keep in touch with the roadmap -- and I can't wait to hear more about your work with TYOM!

best,

-m

0