Michał Karzyński

We tested the solution with `NullableString` and it works well. However it still add fields with null values (`"optional_field": null`) to our models. When we persist these models, we either...

Is there a way to submit updated test data for review?

I would insist that the current package content should be installable from PyPI. However, the package could be split into two and the one package could contain only core functionality,...

Reducing the size of the ONNX package does sound like a good idea. However, there are scenarios which do utilize the package in order to run testing. In particular, the...

I can confirm this issue still persists in 2.7.0

I think we should avoid creating a new version of the operator, which acts a a superset of both behaviours. This would place the burden on implementors, to support both...

This PR changes the intention of the development version on the Scoreboard. If there is no dev version of ONNX Runtime, then we should probably just remove it from the...

I'm guessing Python 3.9 and: ``` torch==1.7.1 scipy==1.5.4 ``` Can be a good starting point...

I have a similar question. It is currently possible to specify custom `requirements.txt` file. I'd like to have the ability to specify a custom `Dockerfile` as part of the task...

I suppose using the [`set_base_docker`](https://clear.ml/docs/latest/docs/references/sdk/task#set_base_docker) method of a task can help address most scenarios.