Views
No views yet
1import segmentation_models_pytorch as smp
2
3model = smp.from_pretrained("<save-directory-or-this-repo>")1model_init_params = {
2 "encoder_name": "efficientnet-b3",
3 "encoder_depth": 5,
4 "encoder_weights": "imagenet",
5 "decoder_use_batchnorm": True,
6 "decoder_channels": (256, 128, 64, 32, 16),
7 "decoder_attention_type": None,
8 "in_channels": 3,
9 "classes": 2,
10 "activation": None,
11 "aux_params": None
12}