Issues with learn_one and the documentation

[hairuoguo]

Hello!

I am currently trying to use this library for a project, but there are a couple of things that have been blocking my progress.

1. The documentation site is down -- would it be possible to have the documentation hosted on another site or to revive the documentation page? It would make it a lot easier to figure out what functions are supposed to do/the intended types and descriptions for certain arguments.
2. I'm currently trying to work with the SoftMax model in linear_models. However, I've been having trouble with the learn_one method for that model. To be more precise, I end up with a "unsupported operand type(s) for *: 'dict' and 'float'" error for line 96 in linear_model/softmax. My x input is a dictionary with str keys for features and float values for feature values (input dim size 1, this is a toy problem), and I've tried formatting the y both as a single int/str class label and as a dictionary of floats (keys as labels, values as desired confidence scores). What would be the appropriate formatting to ensure that the learn/predict functions work properly? (some syntactic sugar where the functions are declared would also be helpful in lieu of complete documentation). Thank you!

[MaxHalford]

Hey there.

The documentation site is down -- would it be possible to have the documentation hosted on another site or to revive the documentation page? It would make it a lot easier to figure out what functions are supposed to do/the intended types and descriptions for certain arguments.

Which site are you talking about? https://riverml.xyz/latest/ is up for me

I'm currently trying to work with the SoftMax model in linear_models. However, I've been having trouble with the learn_one method for that model. To be more precise, I end up with a "unsupported operand type(s) for *: 'dict' and 'float'" error for line 96 in linear_model/softmax. My x input is a dictionary with str keys for features and float values for feature values (input dim size 1, this is a toy problem), and I've tried formatting the y both as a single int/str class label and as a dictionary of floats (keys as labels, values as desired confidence scores). What would be the appropriate formatting to ensure that the learn/predict functions work properly? (some syntactic sugar where the functions are declared would also be helpful in lieu of complete documentation). Thank you!

How about sharing an MRE with some code?

[hairuoguo]

Ah, apparently the documentation site wasn't loading for me due to my being on a VPN, but am able to access it once VPN is turned off. Was able to resolve issue by looking at documentation. Thanks!