AutoMLPreset
- class lightautoml.automl.presets.base.AutoMLPreset(task, timeout=3600, memory_limit=16, cpu_limit=4, gpu_ids='all', debug=False, timing_params=None, config_path=None, **kwargs)[source]
Bases:
AutoML
Basic class for automl preset.
It’s almost like AutoML, but with delayed initialization. Initialization starts on fit, some params are inferred from data. Preset should be defined via
.create_automl
method. Params should be set via yaml config. Most usefull case - end-to-end model development.Commonly _params kwargs (ex. timing_params) set via config file (config_path argument). If you need to change just few params, it’s possible to pass it as dict of dicts, like json. To get available params please look on default config template. Also you can find there param description. To generate config template call
SomePreset.get_config('config_path.yml')
.Example
>>> automl = SomePreset(Task('binary'), timeout=3600) >>> automl.fit_predict(data, roles={'target': 'TARGET'})
- Parameters:
task (
Task
) – Task to solve.timeout (
int
) – Timeout in seconds.memory_limit (
int
) – Memory limit that are passed to each automl.cpu_limit (
int
) – CPU limit that that are passed to each automl.gpu_ids (
Optional
[str
]) – GPU IDs that are passed to each automl.verbose – Controls the verbosity: the higher, the more messages. <1 : messages are not displayed; >=1 : the computation process for layers is displayed; >=2 : the information about folds processing is also displayed; >=3 : the hyperparameters optimization process is also displayed; >=4 : the training process for every algorithm is displayed;
**kwargs (
Any
) – Not used.
- create_automl(**fit_args)[source]
Abstract method - how to build automl.
Here you should create all automl components, like readers, levels, timers, blenders. Method
._initialize
should be called in the end to create automl.- Parameters:
**fit_args – params that are passed to
.fit_predict
method.
- fit_predict(train_data, roles, train_features=None, cv_iter=None, valid_data=None, valid_features=None, verbose=0)[source]
Fit on input data and make prediction on validation part.
- Parameters:
train_data (
Any
) – Dataset to train.roles (
dict
) – Roles dict.train_features (
Optional
[Sequence
[str
]]) – Features names, if can’t be inferred from train_data.cv_iter (
Optional
[Iterable
]) – Custom cv-iterator. For example,TimeSeriesIterator
.valid_features (
Optional
[Sequence
[str
]]) – Optional validation dataset features if can’t be inferred from valid_data.verbose (
int
) – Verbosity level that are passed to each automl.
- Return type:
- Returns:
Dataset with predictions. Call
.data
to get predictions array.
- static set_verbosity_level(verbose)[source]
Verbosity level setter.
- Parameters:
verbose (
int
) – Controls the verbosity: the higher, the more messages. <1 : messages are not displayed; >=1 : the computation process for layers is displayed; >=2 : the information about folds processing is also displayed; >=3 : the hyperparameters optimization process is also displayed; >=4 : the training process for every algorithm is displayed;