TensorFlow Object Detection API —— 制作自己的模型

https://blog.csdn.net/qq_24474463/article/details/81530900

(t20190518) luo@luo-All-Series:~/MyFile/TensorflowProject/Faster_RCNN/models/research$ 

(t20190518) luo@luo-All-Series:~/MyFile/TensorflowProject/Faster_RCNN/models/research$ 

tensorboard  --logdir=training2

 (t20190518) luo@luo-All-Series:~/MyFile/TensorflowProject/Mask_RCNN/mask_rcnn_20190518/Mask_RCNN/models/research$

python ./legacy/eval.py --logtostderr --checkpoint_dir=training2 --pipeline_config_path=training2/ssdlite_mobilenet_v2_coco.config --eval_dir=training2/eval/

========================2019-09-11=======================================

cd /dev/disk_2019/mask_rcnn_20190518/Mask_RCNN_20190902/models/research

python model_main_2019051801.py --pipeline_config_path=object_detection/training20190902/faster_rcnn_resnet101_coco.config --model_dir=object_detection/training20190902 --num_train_steps=50000 --num_eval_steps=10000 --alsologtostderr

cd /dev/disk_2019/mask_rcnn_20190518/Mask_RCNN_20190902/models/research/object_detection
tensorboard --logdir='training20190902'

===============================================================