romeokienzler

romeokienzler

Member Since 7 years ago

CTO and Chief Data Scientist, STSM, IBM CODAIT, Associate Professor for Artificial Intelligence at FH Berne - IBM Academy of Technology, Switzerland

Experience Points
488
follower
Lessons Completed
9
follow
Lessons Completed
29
stars
Best Reply Awards
115
repos

358 contributions in the last year

romeokienzler Most Used Languages
romeokienzler GitHub Stats

115 Pinned

⚡ Project containig related material for my TensorFlow articles
⚡ Sample application from Discover IoT
⚡ info about me
Aug
30
2 weeks ago
push

romeokienzler push IBM/claimed

romeokienzler
romeokienzler

input-url and visualize-timeseries

romeokienzler
romeokienzler

Merge branch 'master' of github.com:IBM/claimed

commit sha: 49f2a858c0bf219c0cc2ee22c0304d8ad0c92eeb

push time in 2 weeks ago
Aug
24
3 weeks ago
Activity icon
issue

romeokienzler issue comment machine-learning-exchange/mlx

romeokienzler
romeokienzler

Add extendable filter categories

Is your feature request related to a problem? Please describe. As of now, MLX can't replicate MAX/DAX as fields in the metadata are missing

Describe the solution you'd like Please add the following meta data field to all asset types:

  • metadata of type JSON

Describe alternatives you've considered Alternatively, metadata for each asset type is static which would make MLX highly inflexible to adjust to the needs of different organizations

romeokienzler
romeokienzler
Activity icon
issue

romeokienzler issue comment machine-learning-exchange/mlx

romeokienzler
romeokienzler

Add extendable filter categories

Is your feature request related to a problem? Please describe. As of now, MLX can't replicate MAX/DAX as fields in the metadata are missing

Describe the solution you'd like Please add the following meta data field to all asset types:

  • metadata of type JSON

Describe alternatives you've considered Alternatively, metadata for each asset type is static which would make MLX highly inflexible to adjust to the needs of different organizations

romeokienzler
romeokienzler
Aug
23
3 weeks ago
pull request

romeokienzler pull request machine-learning-exchange/katalog

romeokienzler
romeokienzler

add input-postgresql.yaml - 1st CLAIMED component

Activity icon
fork

romeokienzler forked machine-learning-exchange/katalog

⚡ MLX Katalog is a project to hold the default content samples to bootstrap Machine Learning eXchange.
romeokienzler Apache License 2.0 Updated
fork time in 3 weeks ago
Aug
19
4 weeks ago
Activity icon
issue

romeokienzler issue elyra-ai/elyra

romeokienzler
romeokienzler

Pipeline Editor: Render ("galaxy inspired") component configuration UI

Is your feature request related to a problem? Please describe. Currently, notebooks only support rendering text fields based on parsed environment variables. Also, KFP pipeline components only support rudimentary parameter descriptions (e.g. type)

This way a complex component like a DL model can't be properly parameterized - especially if parameter visibility is dependent on the choice of other parameters (e.g. kernel type is only a valid option if kernalized regression is chosen)

Describe the solution you'd like Please have a look at this "galaxy tool" - a galaxy tool is a component for the galaxy pipeline tool. In this part of the XML the actual behavior of the property editing UI is expressed (including UI components only visible under certain circumstances)

I'd like to see something similar for Elyra. The configuration could be included in a cell in the notebook so that a citizen data scientist can edit it as well

Either the XML can be used 1:1 or an equivalent can be used. If should be evaluated if the code base from Galaxy could be used (either by proposing the Galaxy team to externalize this functionality so that it becomes a dependency for Galaxy and Elyra) or just copying the code.

Describe alternatives you've considered There are not alternatives

Additional context Example of the component configuration dialog in galaxy image

Activity icon
issue

romeokienzler issue comment elyra-ai/elyra

romeokienzler
romeokienzler

Pipeline Editor: Render ("galaxy inspired") component configuration UI

Is your feature request related to a problem? Please describe. Currently, notebooks only support rendering text fields based on parsed environment variables. Also, KFP pipeline components only support rudimentary parameter descriptions (e.g. type)

This way a complex component like a DL model can't be properly parameterized - especially if parameter visibility is dependent on the choice of other parameters (e.g. kernel type is only a valid option if kernalized regression is chosen)

Describe the solution you'd like Please have a look at this "galaxy tool" - a galaxy tool is a component for the galaxy pipeline tool. In this part of the XML the actual behavior of the property editing UI is expressed (including UI components only visible under certain circumstances)

I'd like to see something similar for Elyra. The configuration could be included in a cell in the notebook so that a citizen data scientist can edit it as well

Either the XML can be used 1:1 or an equivalent can be used. If should be evaluated if the code base from Galaxy could be used (either by proposing the Galaxy team to externalize this functionality so that it becomes a dependency for Galaxy and Elyra) or just copying the code.

Describe alternatives you've considered There are not alternatives

Additional context Example of the component configuration dialog in galaxy image

push

romeokienzler push IBM/claimed

romeokienzler
romeokienzler

commit sha: aa48af7990d8b75471807abba985c12525a6ca8a

push time in 4 weeks ago
Aug
17
1 month ago
Aug
14
1 month ago
Aug
13
1 month ago
Aug
12
1 month ago
push

romeokienzler push romeokienzler/c3

romeokienzler
romeokienzler

add support for parsing description and data type

commit sha: 6ff1149db694ac0a5427b00caada1aad8f6deeff

push time in 1 month ago
Aug
11
1 month ago