File "mtrand.pyx", line 920, in numpy.random.mtrand.RandomState.choice ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (6,) + inhomogeneous part.

[dyhisbig]

训练自己的数据集，出现这个问题 File "mtrand.pyx", line 920, in numpy.random.mtrand.RandomState.choice ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (6,) + inhomogeneous part.

[dyhisbig]

[dyhisbig]

左边是博主的json文件，右边是我的json文件，整体内容还是一样的，不知道是不是labelme版本不一样导致的

[dyhisbig]

应该不是数据集的问题，我下载了shapes数据集博主网盘的数据集，还是会出现上述错误

[dyhisbig]

![Uploading 图片.png…]()

[dyhisbig]

mode = random.choice(self.sample_options) File "mtrand.pyx", line 920, in numpy.random.mtrand.RandomState.choice ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (6,) + inhomogeneous part. 我在kaggle上，找了个数据集，划分好的，还是这个问题，不捣鼓了，没意思

[XYz-dev-max]

博主后来有再了解吗，遇到了相同的问题，没看懂啥意思

[ge548]

mode = random.choice(self.sample_options) File "mtrand.pyx", line 920, in numpy.random.mtrand.RandomState.choice ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (6,) + inhomogeneous part. 我在kaggle上，找了个数据集，划分好的，还是这个问题，不捣鼓了，没意思

augmentations.py中RandomSampleCrop()函数问题，我是注释了这个函数可以运行了，不进行随机裁剪。