mindyolo
mindyolo copied to clipboard
MindSpore YOLO series toolbox and benchmark
如题,请问有无非命令行方法,直接在Python中类似调用API的形式直接进行推理,获得预测值的predict_list这样的使用方法呢?
Thank you for your contribution to the MindYOLO repo. Before submitting this PR, please make sure: - [ ] You have read the [Contributing Guidelines on pull requests](https://github.com/mindspore-lab/mindyolo/blob/master/CONTRIBUTING.md) - [...
Thank you for your contribution to the MindYOLO repo. Before submitting this PR, please make sure: - [ ] You have read the [Contributing Guidelines on pull requests](https://github.com/mindspore-lab/mindyolo/blob/master/CONTRIBUTING.md) - [...
Here is a summary of common training and inference issues.
Thank you for your contribution to the MindYOLO repo. Before submitting this PR, please make sure: - [ ] You have read the [Contributing Guidelines on pull requests](https://github.com/mindspore-lab/mindyolo/blob/master/CONTRIBUTING.md) - [...
For the official training of MindYOLO's YOLOv8, is the hardware device used Ascend 910B or Ascend 910A?
mindyolo是否支持关键点检测、人体姿态估计任务
代码仓:https://github.com/mindspore-lab/mindyolo/tree/v0.3.0 环境:modelart、mindspore_2.2.12-cann_7.0.1.1-py_3.9-euler_2.10.7-aarch64-snt3p、  问题描述: 在使用过程中发现1000+图片、bachsize=64、num_parallel_workers=8时,一个epoch训练时长达到2.24分钟左右,请问如何打印出我还剩多少显存?以及训练一轮2分多时间正常么?若不正常请问有哪些解决方法呢? 请回复下,谢谢! 终端日志如下所示: 
训练使用的yolo格式,test出现这个报错,是因为格式的问题吗,训练测试只能使用coco格式吗
训练期间经常出现Segmentation fault而导致训练中断,每次训练必这样