Closed jimmylihui closed 8 months ago
Hi, @jimmylihui. Thanks for using MogaNet. This repository doesn't provide specific implementations of baselines. We recommend you refer to OpenMixup for our benchmarks of vision backbones on ImageNet, MMDetection and MMSegmentation for the object detection and segmentation tasks, MMPose for the 2D and 3D pose estimation tasks. Another way to benchmark backbone baselines is to train the model name in our training scripts, e.g., training DeiT-S with 8 GPUs as python -m torch.distributed.launch --nproc_per_node=8 --master_port=29501 train.py --model deit_small_patch16_224 --img_size 224 --drop_path 0.1 --batch_size 128 --lr 1e-3 --weight_decay 0.05 --update_freq 1
. This repository mainly provides the pre-training and fine-tuning implementations of MogaNet, while you can simply modify the training script to train the relevant baselines.
Hi there, I will close this issue if there are no more questions. Feel free to contact us if you have new concerns.
Hi, Is there code for training the baseline? Best, Jiahui.Li