Pretrained weights for lucid.models.sk_resnet_34_cls.
Single ImageNet-1k checkpoint (RA_IN1K) converted from
timm/skresnet34.ra_in1k — Wightman's RandAugment recipe, hosted
under huggingface.co/lucid-dl/sk-resnet-34 with the official
acc@1 = 76.956 / acc@5 = 93.320 validation metrics.
Examples
>>> from lucid.models.vision.sknet._weights import SKResNet34Weights
>>> list(SKResNet34Weights.__members__)
['RA_IN1K', 'DEFAULT']
DEFAULT is an alias rather than a fourth entry, so a bare
pretrained=True and the tag it resolves to cannot drift apart.
>>> SKResNet34Weights.DEFAULT is SKResNet34Weights.RA_IN1K
True
>>> SKResNet34Weights.RA_IN1K.num_classes
1000