Object Detection toolkit based on PaddlePaddle. It supports object detection, instance segmentation, multiple object tracking and real-time multi-person keypoint detection.
[X] 我已经搜索过问题,但是没有找到解答。I have searched the question and found no related answer.
请提出你的问题 Please ask your question
The training is not using all the availible resources in the system.
The training commond is: python -u tools/train.py -c configs/ssd/ssdlite_mobilenet_v3_small_320_coco.yml --use_vdl=true --vdl_log_dir=vdl_dir/scalar --eval -o use_gpu=true
Resource used are:
GPU: 12-42%
cpu:: 18-22%
memory: 60%
Disk: 6%
The training eta for 100 epoch is 100. eta is 4 days.
Is there way to set up the training to use more resource to train faster.
Thanks in advance.
问题确认 Search before asking
请提出你的问题 Please ask your question
The training is not using all the availible resources in the system. The training commond is: python -u tools/train.py -c configs/ssd/ssdlite_mobilenet_v3_small_320_coco.yml --use_vdl=true --vdl_log_dir=vdl_dir/scalar --eval -o use_gpu=true Resource used are: GPU: 12-42% cpu:: 18-22% memory: 60% Disk: 6% The training eta for 100 epoch is 100. eta is 4 days. Is there way to set up the training to use more resource to train faster. Thanks in advance.