train.sh

#!/bin/bash
echo "Export params ..."

export HIP_VISIBLE_DEVICES=0,1,2,3 # 自行修改为训练的卡号和数量
export HSA_FORCE_FINE_GRAIN_PCIE=1
export USE_MIOPEN_BATCHNORM=1

echo "Training start ..."
# coco_path是训练数据集地址，数据是coco format
# python -m torch.distributed.launch --nproc_per_node=4 --use_env main.py --coco_path /path/to/coco
# python -m torch.distributed.launch --nproc_per_node=4 --use_env main.py --coco_path /home/datasets/COCO2017
python -m torch.distributed.launch --nproc_per_node=4 --use_env main.py \
	--coco_path /home/datasets/COCO2017 \
	--resume ./pretrained_model/detr-r50-e632da11.pth \
	--lr 1e-6 \
--epochs 5 > log_detr_2_4_dcu.20231205

echo "Training finished."