gnes.encoder.image.onnx module

class gnes.encoder.image.onnx.BaseONNXImageEncoder(model_name: str, model_dir: str, use_cuda: bool = False, *args, **kwargs)[source]

Bases: gnes.encoder.base.BaseImageEncoder

batch_size = 64
encode(img: List[np.ndarray], *args, **kwargs) → numpy.ndarray[source]
post_init()[source]

Declare class attributes/members that can not be serialized in standard way

train(*args, **kwargs)

Train the model, need to be overrided