hydra | agnostic ML Platform that will enable Data Scientists | Machine Learning library

The documentation in https://hydra.cc/docs/patterns/select_multiple_configs_from_config_group/ shows how to pick multiple configs from a Config Group and place them in a dictionary-like structure.

However, as mentioned in the very last paragraph there, "example uses an explicit nesting level inside each of the configs to prevent them stepping over one another". For my use-case, this would prove extremely cumbersome and I would like to avoid it at all costs if possible.

Is there a way to achieve a similar result without resorting to explicitly adding the level in the individual configs? Thanks in advance :)

After upgrading to Api platform 2.6, the non ApiResource entities started being serialized (ld-json) along with a fake identifier and their type. For example:

Hydra provides a way to dynamically create a hierarchical configuration by composition and override it through config files and the command line, leveraging OmegaConf. I look for a recommended way to document the parameters but I could not find (a documented) one. What are best practices for that? Coming from argparse, I like the way of documenting the parameter inline, i.e. close to the code.

Please, see this question for context and the answer by MrC aka Shaun Curtis

This question is about something I've tried to solve in the past without great success. Are you familiar with this sample I've once downloaded and run it. It did not work. I then realized that I must add the base url to the url in the browser's address bar in order to run the first project, for instance: https://localhost: 44302/FirstApp That is, the Client project. And for the SecondClient it should be https://localhost: 44302/SecondApp. This is exactly how the sample app by MrC aka Shaun Curtis works, though he added a Razor Pages app to provide a menu for redirection to the four projects.

What I tried to do without much success is to make the first WebAssemby front end project, which is hosted, to be the default; that is when I run the app, or type in the address bar https://localhost: 44302. And if I type https://localhost: 44302/FirstApp I see the first stand alone WebAssembly project I added to the solution. And a second project, and a third project, and so on, all are WebAssembly projects. I could not do that: When I ran the default project everything is fine... I can navigate within the bounds of the project, route to the Counter page, FetchData page, etc.

But when I add the segment /FirstApp to the url in the address bar and hit enter, the Router displays the message "Sorry, there's nothing at this address." instead of navigating to the project represented by the base url /FirstApp/

Does anyone here have any idea how to achieve the requested feature I am looking for ?

Recently I have started to use hydra to manage the configs in my application. I use Structured Configs to create schema for .yaml config files. Structured Configs in Hyda uses dataclasses for type checking. However, I also want to use some kind of validators for some of the parameter I specify in my Structured Configs (something like this).

Do you know if it is somehow possible to use Pydantic for this purpose? When I try to use Pydantic, OmegaConf complains about it:

In Hydra I have the following configuration:

I have a project where I want to build a full-blown IDP (using Golang). So technically, a user wants to SSO into another system using my service. I am looking to build this service from scratch. Upon researching for open-source IDP solutions, I came across ory/Hydra and ory/Kratos. I went through their documentation and did a quick-start tutorial. I am still confused about which of the above 2 libraries are suitable for the development of this service.

From a high-level standpoint, this is what I am trying to do.

1. There is an existing third-party web app X.
2. A user has already login credentials for my service Y.
3. A user wants to post some comments in app X.
4. He/She is redirected to my service Y login screen if not logged in.
5. After login, the Callback URL returns him back to app X comment section.

Is there a better way to reload a hydra config from an experiment with enumerations? Right now I reload it like so:

I would like to use python datatypes - both built-in and imported from libraries such as numpy, tensorflow, etc - as arguments in my hydra configuration. Something like: