Typo in "How to Get Started" Documentation

[will-pang]

Fixing a small typo in the "how_to_get_started" documentation.

I believe for this eval/monitoring method:

trainer.train(
    train_dataloader=train_loader,
    val_dataloader=val_loader,
    epochs=50,
    monitor="pr_auc_samples",
)

We want monitor=pr_auc as prc_auc_samples is for multi label classification, whereas pr_auc is for binary classification (yes/no decease).

Running the current documentation leads the pipeline to error out/fail.

[jhnwu3]

Ah thanks for the catch William! That's on us , we also need to sooner or later update all of these installation instructions as we're aiming to get a full PyHealth 2.0 release later this next week. We're super close to basically revamping the entire package to work with our new backend.