Can Rapidminer save a machine learning model in ONNX format?
Hello,
I have been reading about the open format for machine learning models called ONNX - Open Neural Network Exchange. See onnx.ai for more information.
Is it curently possible to save RapidMiner models in this format? If not, is that something that is considered for a future release?
Thanks very much,
Jeroen
Find more posts tagged with
Sort by:
1 - 4 of
41
MartinLiebig
Altair EmployeeHi,
we don't support it right now.
Best,
Martin
New Altair Community Member
OPHi Martin,
Thanks for the reply. One question remains: is it something that is considered for a future release or is it just never going to happen?
Regards,
Jeroen
New Altair Community Member
Hi Jeroen,
Many RapidMiner-generated models have embedded information that is used for the Simulator and other features that are part of the RapidMiner platform. Exporting those nuances to the ONNX format can be a little difficult. Take this with a grain of salt, though, as it is my opinion as a user and not part of the RapidMiner core.
While you cannot export your RapidMiner models to ONNX, there is something you can do though, if you are in need for ONNX-RapidMiner interaction: you can use the Python extension to work with ONNX models inside RapidMiner, which internally hooks RapidMiner datasets into Pandas dataframes to be used inside a script.
All the best,
Rodrigo.
Many RapidMiner-generated models have embedded information that is used for the Simulator and other features that are part of the RapidMiner platform. Exporting those nuances to the ONNX format can be a little difficult. Take this with a grain of salt, though, as it is my opinion as a user and not part of the RapidMiner core.
While you cannot export your RapidMiner models to ONNX, there is something you can do though, if you are in need for ONNX-RapidMiner interaction: you can use the Python extension to work with ONNX models inside RapidMiner, which internally hooks RapidMiner datasets into Pandas dataframes to be used inside a script.
All the best,
Rodrigo.