ResNetModel
- class lightautoml.ml_algo.torch_based.nn_models.ResNetModel(n_in, n_out=1, hid_factor=[2, 2], drop_rate=0.1, noise_std=0.05, act_fun=torch.nn.ReLU, num_init_features=None, use_bn=True, use_noise=False, device=torch.device, **kwargs)[source]
Bases:
Module
The ResNet model from https://github.com/Yura52/rtdl.
- Parameters:
n_in (
int
) – Input dim.n_out (
int
) – Output dim.hid_factor (
List
[float
]) – Dim of intermediate fc is increased times this factor in ResnetModel layer.drop_rate (
Union
[float
,List
[float
],List
[List
[float
]]]) – Dropout rate for each layer separately or altogether.noise_std (
float
) – Std of noise.act_fun (
Module
) – Activation function.num_init_features (
Optional
[int
]) – If not none add fc layer before model with certain dim.use_bn (
bool
) – Use BatchNorm.use_noise (
bool
) – Use noise.device (
device
) – Device to compute on.