syne_tune.optimizer.schedulers.searchers.bayesopt.models.model_skipopt module
- class syne_tune.optimizer.schedulers.searchers.bayesopt.models.model_skipopt.SkipOptimizationPredicate[source]
Bases:
object
Interface for
skip_optimization
predicate inModelStateTransformer
.
- class syne_tune.optimizer.schedulers.searchers.bayesopt.models.model_skipopt.NeverSkipPredicate[source]
Bases:
SkipOptimizationPredicate
Hyperparameter optimization is never skipped.
- class syne_tune.optimizer.schedulers.searchers.bayesopt.models.model_skipopt.AlwaysSkipPredicate[source]
Bases:
SkipOptimizationPredicate
Hyperparameter optimization is always skipped.
- class syne_tune.optimizer.schedulers.searchers.bayesopt.models.model_skipopt.SkipPeriodicallyPredicate(init_length, period, metric_name='target')[source]
Bases:
SkipOptimizationPredicate
Let
N
be the number of labeled points for metricmetric_name
. Optimizations are not skipped ifN < init_length
. Afterwards, we increase a counter wheneverN
is larger than in the previous call. With respect to this counter, optimizations are done everyperiod
times, in between they are skipped.- Parameters:
init_length (
int
) – See aboveperiod (
int
) – See abovemetric_name (
str
) – Name of internal metric. Defaults toINTERNAL_METRIC_NAME
.
- class syne_tune.optimizer.schedulers.searchers.bayesopt.models.model_skipopt.SkipNoMaxResourcePredicate(init_length, max_resource, metric_name='target')[source]
Bases:
SkipOptimizationPredicate
This predicate works for multi-fidelity HPO, see for example
GPMultiFidelitySearcher
.We track the number of labeled datapoints at resource level
max_resource
. HP optimization is skipped if the total numberN
of labeled cases isN >= init_length
, and if the number ofmax_resource
cases has not increased since the last recent optimization.This means that as long as the dataset only grows w.r.t. cases at lower resources than
max_resource
, this does not trigger HP optimization.- Parameters:
init_length (
int
) – See abovemax_resource (
int
) – See abovemetric_name (
str
) – Name of internal metric. Defaults toINTERNAL_METRIC_NAME
.