Skip to content

Milestones

List view

  • This milestone tracks our revamp of the model training and eval pipelines. I plan to use modal to automate these, so that it's much easier to retrain and evaluate models as my dataset changes. I want to focus on updating the data, instead the models themselves, and create the highest quality datasets possible.

    Overdue by 1 year(s)
    Due by August 24, 2024
    0/1 issues closed
  • Right now, we have a feature to conduct the optimal ev calcs. Shouldn't be too hard to train a model that can do damage calcs in general.

    No due date
  • Speedchecks feature currently only uses a dict to parse inputs. I want to make it use my Pokemon and Gamestate classes so that this makes more sense with how the optimal ev calcs works.

    No due date