from_bytes¶

Load a model from the binary format written by to_bytes. This is a class method.

Method call format¶

TBoostRegressor.from_bytes(data)

Parameters¶

data¶

Description¶

The output of to_bytes.

Possible types

bytes

Default value

Required parameter

Return value¶

A new, trained TBoostRegressor.

t-boost reads the files written by the same or an earlier version, and a loaded model predicts identically to the version that wrote it. There are two exceptions: a file written by a newer version raises SerializationError naming both versions, and a model with categorical features saved by t-boost 0.6 is refused, because its missing level was labelled differently. See Saving and loading models.

Usage examples¶

from t_boost import TBoostRegressor

with open("model.tboost", "rb") as f:
    model = TBoostRegressor.from_bytes(f.read())