mindformers.pet.pet_config.PetConfig

class mindformers.pet.pet_config.PetConfig(pet_type: str = None, **kwargs)[source]

The configuration base class for Parameter-Efficient Tuning (Pet) algorithms.

Parameters: pet_type (str, optional) – The Pet method type. Default: None.
Returns: An instance of PetConfig.

Examples

>>> from mindformers.pet.pet_config import LoraConfig
>>> config = LoraConfig(target_modules='.*wq|.*wk|.*wv|.*wo')
>>> print(config)
{'pet_type': 'lora', 'lora_rank': 8, 'lora_alpha': 16,
'lora_dropout': 0.01, 'lora_a_init': 'normal', 'lora_b_init'
: 'zero', 'param_init_type': mindspore.float16, 'compute_dtype':
mindspore.float16, 'target_modules': '.*wq|.*wk|.*wv|.*wo', 'exclude_layers': None
, 'freeze_include': None, 'freeze_exclude': None}