JL_Cockpit.training_loop — Functiontraining_loop(;
model = nothing, dataset_train = nothing, dataset_test = nothing,
batchsize = 128, epochs = 5, optim = nothing,
vis_loss::Bool = true, vis_grad_norm::Bool = true,
vis_hist_1d::Bool = true, vis_params::Bool = true,
vis_distance::Bool = true, vis_update_size::Bool = true, vis_hist_2d::Bool = false
)Train with AD and visualise live metrics.
JL_Cockpit.visualiser — Functionvisualiser(;
vis_loss::Bool = true, vis_grad_norm::Bool = true, vis_hist_1d::Bool = true,
vis_params::Bool = true, vis_distance::Bool = true, vis_update_size::Bool = true,
vis_hist_2d::Bool = false
)Initialises the visualiser. It will take the given Observables and display a plot that updates live.