transformer_big¶

lucid.models.transformer_big(src_vocab_size: int = 12000, tgt_vocab_size: int = 12000) → Transformer¶

The transformer_big function creates a larger variant of the Transformer class, configured for more complex tasks requiring increased model capacity.

Total Parameters: 213,237,472

Function Signature¶

@register_model
def transformer_big(src_vocab_size: int = 12000, tgt_vocab_size: int = 12000) -> Transformer

Parameters¶

src_vocab_size (int, optional): Size of the source vocabulary. Default is 12,000.
tgt_vocab_size (int, optional): Size of the target vocabulary. Default is 12,000.

Returns¶

Transformer: An instance of Transformer with larger capacity for improved performance on complex tasks.

Examples¶

>>> import lucid.models as models
>>> transformer = models.transformer_big()
>>> print(transformer)
Transformer(src_vocab_size=12000, tgt_vocab_size=12000, d_model=1024, ...)