无法在ios (El Capitan)中安装mxnet并运行python (2.7)示例

Question

我曾尝试运行install Mxnet deeplearning framework，但由于在尝试运行python示例时出现以下错误而失败，并且我在web上找不到最佳的处理方法。

MacBook-Pro-4:mxnet chinhiroshi$ python example/image-classification/train_mnist.py
2016-03-22 11:51:59,386 Node[0] start with arguments Namespace(batch_size=128, data_dir='mnist/', gpus=None, kv_store='local', load_epoch=None, lr=0.1, lr_factor=1, lr_factor_epoch=1, model_prefix=None, network='mlp', num_epochs=10, num_examples=60000)
Traceback (most recent call last):
  File "example/image-classification/train_mnist.py", line 130, in <module>
    train_model.fit(args, net, get_iterator(data_shape))
  File "/Users/chinhiroshi/Dropbox/code/python/package/mxnet/example/image-classification/train_model.py", line 41, in fit
    save_model_prefix = args.save_model_prefix
AttributeError: 'Namespace' object has no attribute 'save_model_prefix'

Answer 1

save_model_prefix是15天前才推出的。您很可能在此之前安装了mxnet，因此有一个不支持它的版本。

您可以从源代码重新安装最新的mxnet，也可以修复示例本身中的问题。对于该打开的train_model.py (请参见回溯中的完整路径)，并更改以下四行：

save_model_prefix = args.save_model_prefix
if save_model_prefix is None:
    save_model_prefix = model_prefix
checkpoint = None if save_model_prefix is None else mx.callback.do_checkpoint(save_model_prefix)

通过以下方式：

checkpoint = None if model_prefix is None else mx.callback.do_checkpoint(model_prefix)