Links
Comment on page

AI Experiment Object

An AI Experiment is a grouping of SigOpt Runs where different parameter configurations yield different metric values.
Key
Type
Value
client
string
The id of the Client who created the SigOpt AI Experiment.
conditionals
created
int
The timestamp (in seconds since epoch) that this experiment was created.
id
string
A unique ID for a given experiment.
linear_constraints
metadata
map<string, value>
Optional user-provided metadata.
metrics
array<Metric>
See Metrics for more information.
name
string
A user-specified name for this experiment.
num_solutions
int
See Multisolution for more information.
budget
int
See Budget for more information.
parallel_bandwidth
int
The number of simultaneous Runs you plan to maintain during this experiment. The default value for this is 1, i.e., a sequential experiment. The maximum value for this is dependent on your plan. This field is optional, but setting it correctly may improve performance.
parameters
array<Parameter>
An array of parameter objects.
progress
Progress
The current progress of this experiment.
project
string
The id of the project that a given experiment belongs to.
state
string
The state of this experiment. Can be active (for experiments that are currently running), or deleted (for experiments that have been deleted).
type
string
A type for this experiment. Used for random and grid.
updated
int
The timestamp (in seconds since epoch) that this experiment was last updated.
user
string
The id of the user who created the experiment.