4_Training.md

Training a model

You could optionally add extra command line parameters --batch_size ${BATCH_SIZE} and --epochs ${EPOCHS} to specify your preferred parameters.

1. Train using Multi-GPUs

• Train with multiple GPUs or multiple machines

sh scripts/dist_train.sh ${NUM_GPUS} --cfg_file ${CONFIG_FILE} --batch_size ${BATCH_NUM}
# or
sh scripts/torch_train.sh ${NUM_GPUS} --cfg_file ${CONFIG_FILE} --batch_size ${BATCH_NUM}
# or
sh scripts/slurm_train.sh ${PARTITION} ${JOB_NAME} ${NUM_GPUS} --cfg_file ${CONFIG_FILE}

• If you use pytorch 1.x, you have to use python -m torch.distributed.launch i.e., tools/scripts/dist_X.sh
• If you use pytorch 2.x, you have to use torchrun i.e., tools/scripts/torch_train_X.sh

cd ~/CenterPointPillar
# you can link as `output` directory from `/Dataset/Train_Results/CenterPoint/` 
ln -s /Dataset/Train_Results/CenterPoint/ output   
cd tools/
sh scripts/torch_train.sh 2 --cfg_file ./cfgs/waymo_models/centerpoint_pillar_train.yaml --batch_size 24

2. Train using Single-GPU

• Train with a single GPU:

python train.py --cfg_file ${CONFIG_FILE}

cd ~/CenterPointPillar
# you can link as `output` directory from `/Dataset/Train_Results/CenterPoint/`
ln -s /Dataset/Train_Results/CenterPoint/ output
cd tools/
CUDA_VISIBLE_DEVICES=1 python train.py --cfg_file ./cfgs/waymo_models/centerpoint_pillar_train.yaml --batch_size 16  # you can replace `CUDA_VISIBLE_DEVICES=1` with gpu's number you want

3. Pretrained Models

• If you would like to train CaDDN, download the pretrained DeepLabV3 model and place within the checkpoints directory.
• Please make sure the kornia is installed since it is needed for CaDDN.

CenterPointPillar
├── checkpoints
│   ├── deeplabv3_resnet101_coco-586e9e4e.pth
├── data
├── pcdet
├── tools

Return to the main page.