PulseFocusPlatform/static/docs/tutorials/train_on_kunlun.md

How to train on kunlun

Prepare kunlun environment

Paddle installation for machines with Kunlun XPU card

yolov3

Prepare data

Prepare data roadsign:

cd PaddleDetection/static/dataset/roadsign_voc/
python3.7 download_roadsign_voc.py

Train

python3.7 -u tools/train.py -c configs/yolov3_mobilenet_v1_roadsign.yml -o use_gpu=False use_xpu=True

Eval

python3.7 -u tools/eval.py -c configs/yolov3_mobilenet_v1_roadsign.yml -o weights=output/yolov3_mobilenet_v1_roadsign/model_final.pdparams use_gpu=False use_xpu=True

Train on Darknet

cd static/
python3.7 -u tools/train.py -c configs/yolov3_datknet_roadsign_kunlun.yml -o use_gpu=False use_xpu=True

Eval on Darknet

cd static/
python3.7 -u tools/eval.py -c configs/yolov3_darknet_roadsign_kunlun.yml -o weights=output/yolov3_darknet_roadsign_kunlun/model_final.pdparams use_gpu=False use_xpu=True

ppyolo

Prepare data

Prepare data roadsign

Train

python3.7 -u tools/train.py --eval -c configs/ppyolo/ppyolo_roadsign_kunlun.yml

Eval

python3.7 -u tools/eval.py -c configs/ppyolo/ppyolo_roadsign_kunlun.yml

mask_rcnn

Prepare data

Download dataset from https://dataset.bj.bcebos.com/PaddleDetection_demo/cocome.tar and put it in the dataset directory.

Train

python3.7 -u tools/train.py --eval -c configs/mask_rcnn_r50_1x_cocome_kunlun.yml

Eval

python3.7 -u tools/eval.py -c configs/mask_rcnn_r50_1x_cocome_kunlun.yml