A Model Tier is what controls the threshold values at which models change accuracy and data per kill values.
Infinitely many model tiers may be created, allowing total control over the progression system.
This object references the following objects:
{
"required_data": integer, // [Mandatory] || The required data threshold for the model tier. Can be overridden by individual data models.
"data_per_kill": integer, // [Mandatory] || The amount of data received for each mob kill while using the deep learner. Can be overridden by individual data models.
"color": Color, // [Mandatory] || The color of the tier, used when displaying the tier in text form.
"accuracy": float, // [Mandatory] || The accuracy (chance a prediction is created) when simulating the model. A value of 1.0 is 100%, and the max value is 64.
"can_sim": boolean // [Optional] || If the model tier is legal for use in a simulation chamber. Defaults to true.
}Model Tiers may only be created under the hostilenetworks namespace; attempting to register a tier under a different namespace will error.
As such, tiers are not namespaced when being referenced, as the namespace is implied to be hostilenetworks.
A tier with a required_data of zero must always exist. The game will refuse to load if no such tier exists.
No two tiers with the same required_data may exist. A linear order of tiers must be formed for the data set to be considered legal.
When creating a new tier, you will also need to supply a language key of the form hostilenetworks.tier.name for translation.
The Superior model tier.
{
"required_data": 354,
"data_per_kill": 18,
"color": "light_purple",
"accuracy": 0.65
}