Source code for autogluon.tabular.models.text_prediction.text_prediction_v1_model
from __future__ import annotations
import logging
from autogluon.common.features.types import (
R_CATEGORY,
R_FLOAT,
R_INT,
R_OBJECT,
S_IMAGE_PATH,
S_TEXT_AS_CATEGORY,
S_TEXT_NGRAM,
S_TEXT_SPECIAL,
)
from ..automm.automm_model import MultiModalPredictorModel
logger = logging.getLogger(__name__)
[docs]
class TextPredictorModel(MultiModalPredictorModel):
"""MultimodalPredictor that doesn't use image features"""
ag_key = "AG_TEXT_NN"
ag_name = "TextPredictor"
_supported_problem_types = ["binary", "multiclass", "regression"]
_default_auxiliary_params_extra = dict(
valid_raw_types=[R_INT, R_FLOAT, R_CATEGORY, R_OBJECT],
ignored_type_group_special=[S_TEXT_NGRAM, S_TEXT_AS_CATEGORY, S_TEXT_SPECIAL, S_IMAGE_PATH],
)