Пример #1
0
def test_init(initialize_bart, initialize_t5):
    bart_tokenizer, _ = initialize_bart
    t5_tokenizer, _ = initialize_t5
    _init(bart_tokenizer, 'facebook/bart-base')
    _init(t5_tokenizer, 't5-base')

    with pytest.raises(ValueError):
        te.SplitterEncoder("not-supported-model")
Пример #2
0
def initialize_t5():
    from transformers import T5Tokenizer
    t5_tokenizer = T5Tokenizer.from_pretrained('t5-base')
    t5_splitter = te.SplitterEncoder('t5-base')
    return t5_tokenizer, t5_splitter
Пример #3
0
def _init(tokenizer, model_name):
    splitter = te.SplitterEncoder(model_name)
    assert type(splitter.tokenizer) is type(tokenizer)