mmpretrain/configs/mlp_mixer/mlp-mixer-large-p16_64xb64_in1k.py

_base_ = [
    '../_base_/models/mlp_mixer_large_patch16.py',
    '../_base_/datasets/imagenet_bs64_mixer_224.py',
    '../_base_/schedules/imagenet_bs4096_AdamW.py',
    '../_base_/default_runtime.py',
]

optim_wrapper = dict(clip_grad=dict(max_norm=1.0))