Can't find ResNext in vision.models

Ups, then maybe this can help you: Using Pretrained Models from Cadene repository