search

mit-han-lab / efficientvit

EfficientViT is a new family of vision models for efficient high-resolution vision.
Apache License 2.0
1.6k stars 142 forks source link

Please show train cls model yaml of l2 r288 and l2 r384, Thanks #46

Closed jiujiuwei closed 8 months ago

jiujiuwei commented 8 months ago

Please show train cls model yaml of l2 r288 and l2 r384 Thanks

han-cai commented 8 months ago

The training commands for l2-r288 and l2-r384 are attached below:

torchpack dist-run -np 16 -H $server1:8,$server2:8 \
python train_cls_model.py configs/cls/imagenet/l2.yaml --fp16 \
    --data_provider.image_size "[128,160,192,224,256,288]" \
    --run_config.eval_image_size "[288]" \
    --path .exp/cls/imagenet/l2_r288/
torchpack dist-run -np 16 -H $server1:8,$server2:8 \
python train_cls_model.py configs/cls/imagenet/l2.yaml --fp16 \
    --data_provider.image_size "[128,160,192,224,256,288,320,352,384]" \
    --run_config.eval_image_size "[384]" \
    --path .exp/cls/imagenet/l2_r384/