In this documentation, we describe more details about training PPlanedet.
If you only have one gpu, you can run the following codes
# training on single-GPU
export CUDA_VISIBLE_DEVICES=0
python tools/train.py -c configs/scnn/resnet50_tusimple.py # please change to your config path
If you have more than one gpu, you can run the following codes
# training on 4 GPUs
export CUDA_VISIBLE_DEVICES=0,1,2,3
python -m paddle.distributed.launch tools/train.py -c configs/scnn/resnet50_tusimple.py # please change to your config path
If you want to load the pretrain weight, you can run
# training on single-GPU
export CUDA_VISIBLE_DEVICES=0
python tools/train.py -c configs/scnn/resnet50_tusimple.py --load pretrain weight path #
If you want to resume the training program, you can run
# training on single-GPU
export CUDA_VISIBLE_DEVICES=0
python tools/train.py -c configs/scnn/resnet50_tusimple.py --resume weight path #