guidelines to implement custom models

[MiqG]

Hi, thanks for developing such a comprehensive tool for using sequence models.

I was wondering where I could find the set of methods I should include to be able to use your functions on a model that is still under development.

I understand you download your models from a weights and biases instance. But, what are the methods and naming conventions you had to include in the models such that can be run with your framework seamlessly?

Thanks very much in advance! This is a very helpful tool! Best,

Miquel

[avantikalal]

Hi @MiqG, currently it is not possible to use a custom model directly, but we are working to enable this and it will be added soon. Basically, the custom model class would have to inherit from our BaseModel class (https://github.com/Genentech/gReLU/blob/main/src/grelu/model/models.py#L24).