Plan Rules Object
Last updated
Last updated
Key | Type | Value |
---|---|---|
max_categorical_breadth
int
How many distinct categorical values experiments can have across all parameters.
max_checkpoints
int
The max number of training monitor checkpoints that can be created during a training run.
max_clients
int
How many individual clients (teams) can use the product.
max_conditionals_breadth
int
How many distinct conditional values experiments can have across all parameters.
max_constraints
int
How many parameter constraints experiments can have.
max_dimension
int
How many parameters experiments can have.
max_experiments
int
How many experiments the client is allowed to create in each period.
max_extra_failures
int
How many failed observations are allowed to occur.
max_metrics
int
How many metrics can be optimized at once. Does not include unoptimized metrics.
max_observations
int
How many observations each experiment can have.
max_parallelism
int
How many open suggestions each experiment can have concurrently.
max_solutions
int
How many solutions experiments can have.
max_users
int
How many users the organization can have.
supports_benchmarking
boolean
Whether creating random and grid experiments is allowed.
supports_multitask
boolean
Whether multitask experiments are allowed.