SquadTensorizer.ConfigΒΆ

Component: SquadTensorizer

class SquadTensorizer.Config[source]

Bases: TokenTensorizer.Config

All Attributes (including base classes)

is_input: bool = True
column: str = 'text'
tokenizer: Tokenizer.Config = Tokenizer.Config(split_regex='\\W+')
add_bos_token: bool = False
add_eos_token: bool = False
use_eos_token_for_bos: bool = False
max_seq_len: Optional[int] = None
vocab: VocabConfig = VocabConfig()
vocab_file_delimiter: str = ' '
doc_column: str = 'doc'
ques_column: str = 'question'
answers_column: str = 'answers'
answer_starts_column: str = 'answer_starts'
max_ques_seq_len: int = 64
max_doc_seq_len: int = 256
Subclasses
  • SquadTensorizerForKD.Config

Default JSON

{
    "is_input": true,
    "column": "text",
    "tokenizer": {
        "Tokenizer": {
            "split_regex": "\\W+",
            "lowercase": true,
            "use_byte_offsets": false
        }
    },
    "add_bos_token": false,
    "add_eos_token": false,
    "use_eos_token_for_bos": false,
    "max_seq_len": null,
    "vocab": {
        "build_from_data": true,
        "size_from_data": 0,
        "min_counts": 0,
        "vocab_files": []
    },
    "vocab_file_delimiter": " ",
    "doc_column": "doc",
    "ques_column": "question",
    "answers_column": "answers",
    "answer_starts_column": "answer_starts",
    "max_ques_seq_len": 64,
    "max_doc_seq_len": 256
}