https://blog.youkuaiyun.com/w05072zh-cnhttps://blog.youkuaiyun.com/https://blog.youkuaiyun.com/w05072/article/details/139454492https://blog.youkuaiyun.com/w05072/article/details/139454492w05072Tue, 04 Jun 2024 21:28:56 +0800https://blog.youkuaiyun.com/w05072/article/details/139248699https://blog.youkuaiyun.com/w05072/article/details/139248699w05072Mon, 27 May 2024 21:44:45 +0800https://blog.youkuaiyun.com/w05072/article/details/136585796https://blog.youkuaiyun.com/w05072/article/details/136585796w05072Sat, 09 Mar 2024 16:18:24 +0800= 0 && t < n_classes` failed，然后我百度了一下发现是损失函数哪里报错了，他有其他除了255和类别之外的标签值，然后我确认了所有训练集图片，发现都是对的。然后读取时也使用numpy.unique（label）发现没有问题，但是在minibatch加载时出现了254这个标签值，百思不得其解，然后！然后我发现确实是我没有读取到标签转换后的路径，然后改完路径之后发现，又又又又出现了这个问题，我人都懵了，一直找不到答案。]]>https://blog.youkuaiyun.com/w05072/article/details/135895831https://blog.youkuaiyun.com/w05072/article/details/135895831w05072Sun, 28 Jan 2024 15:36:23 +0800https://blog.youkuaiyun.com/w05072/article/details/135460257https://blog.youkuaiyun.com/w05072/article/details/135460257w05072Mon, 08 Jan 2024 16:20:23 +0800https://blog.youkuaiyun.com/w05072/article/details/135372862https://blog.youkuaiyun.com/w05072/article/details/135372862w05072Wed, 03 Jan 2024 20:51:18 +0800