mindformers.models.swin.SwinForImageClassification

class mindformers.models.swin.SwinForImageClassification(config: mindformers.models.base_config.BaseConfig = None)

Swin Transformer Model. The supported model name could be selected from SwinForImageClassification.show_support_list().

参数

config (SwinConfig) – the config of Swin model.

实际案例

>>> # input model name, load model and weights
>>> model_a = SwinForImageClassification.from_pretrained('swin_base_p4w7')
>>> # input config, load model without weights
>>> from mindformers import AutoConfig
>>> config = AutoConfig.from_pretrained('swin_base_p4w7')
>>> model_b = SwinForImageClassification(config)

load_checkpoint(config)

load checkpoint for SwinForImageClassification.

参数

• config (ModelConfig) – a model config instance, which could have attribute

• "checkpoint_name_or_path (str) –

• name or a path to checkpoint, to load model weights. (model) –

remap_pretrained_keys_swin(checkpoint_model)

remap pretrained keys swin