ModelInput

class pytext.models.seq_models.seqnn.ModelInput

Bases: ModelInput

All Attributes (including base classes)

tokens: SeqTokenTensorizer.Config = SeqTokenTensorizer.Config()

dense: Optional[FloatListTensorizer.Config] = None

labels: LabelTensorizer.Config = LabelTensorizer.Config()

Default JSON

{
    "tokens": {
        "is_input": true,
        "column": "text_seq",
        "max_seq_len": null,
        "add_bos_token": false,
        "add_eos_token": false,
        "use_eos_token_for_bos": false,
        "add_bol_token": false,
        "add_eol_token": false,
        "use_eol_token_for_bol": false,
        "tokenizer": {
            "Tokenizer": {
                "split_regex": "\\s+",
                "lowercase": true,
                "use_byte_offsets": false
            }
        },
        "max_turn": 50
    },
    "dense": null,
    "labels": {
        "LabelTensorizer": {
            "is_input": false,
            "column": "label",
            "allow_unknown": false,
            "pad_in_vocab": false,
            "label_vocab": null,
            "label_vocab_file": null,
            "add_labels": null
        }
    }
}