The @dataclass decorator can be set as frozen to prevent modification of
values on instances of that dataclass.
Without making it frozen, I can easily subvert validations by changing the value
of attributes after the __post_init__ validations are called.
>>> config = BPEConfig(300, []) # passes validations
>>> config.vocab_size = 22 # this is invalid, wish this was preventedHere is the updated @dataclass declaration with frozen=True passed as a
parameter.
from dataclasses import dataclass
from typing import ClassVar
@dataclass(frozen=True)
class BPEConfig:
BASE_VOCAB_SIZE: ClassVar[int] = 256
vocab_size: int
special_tokens: list[str]
def __post_init__(self):
if self.vocab_size < self.BASE_VOCAB_SIZE:
msg = f"vocab_size ({self.vocab_size}) must be greater than or equal to BASE_VOCAB_SIZE ({self.BASE_VOCAB_SIZE})"
raise ValueError(msg)Now I am prevented from modifying a scalar value like vocab_size after the
instance has been created.
>>> config = BPEConfig(300, [])
>>> config.vocab_size = 22
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "<string>", line 4, in __setattr__
dataclasses.FrozenInstanceError: cannot assign to field 'vocab_size'This doesn't prevent you from modifying the contents of attributes that are
list or dict types.