猫猫与橙子的博客

1.mxnet报错：Warning: loading symbol saved by MXNet version 10300 with lower version of MXNet v10201. May cause undefined behavior. Please update MXNet if you encounter any issue解决方法：更新mxnet先查看自己...

1.No conversion function registered for op type BlockGrad yet #12682解决方法：>> pip install mxnet --pre参考链接：https://github.com/apache/incubator-mxnet/issues/126822.Unrecognized attribute: s...

pytorch中的网络结构怎样可视化：方法1：使用model_path = '/home/fuxueping/sdb/PycharmProjects/mxnetTopytorch/se_best/model'epoch = 29#model_path路径/home/fuxueping/sdb/PycharmProjects/mxnetTopytorch/se_best中保存的mxne...

转换基本流程：1）创建pytorch的网络结构模型；2）利用mxnet来读取其存储的预训练模型，用于读取模型的参数；3）遍历mxnet加载的模型参数；4）对一些指定的key值，需要进行相应的处理和转换；5）对修改后的层名（key值），利用numpy之间的转换来实现加载；6）对相应层进行参数（feature）进行比较；流程基本是与caffe模型转pytorch模型这篇文章...