Training your own MNASNET (or not)

Image for post
Image for post

Motivation

Image for post
Image for post

What we tried and learned

Image for post
Image for post

What to watch out for

Reuse our code!

self.cluster_dict = { 0: (384,512), 1: (512, 512), 2: (512, 384)}

Typical launch code

CUDA_VISIBLE_DEVICES=0,2 python3 train.py \ --epochs 1000 --epoch_fraction 0.2 \ --batch-size 200 --workers 6 \ --dataset imagenet \ --size_ratio 0.5 --preprocessing_type 5 \ --fold 0 --arch mnasnet --multi_class False \ --num_classes 1000 --classifier_config 512 \ --lr 1e-3 --optimizer adam --lr_regime auto_decay \ --epochs_grow_size 0 \ --lognumber mnasnet_standard_adam_1e3_512_cut_last_nosampler \ --tensorboard True --tensorboard_images True --print-freq 5 \

Final remarks

Data Scientist

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store