keras预训练模型下载地址

keras预训练模型下载地址

这些地址都是从keras文件中找的，总结一下。

vgg16
https://github.com/fchollet/deep-learning-models/releases/download/v0.1/vgg16_weights_tf_dim_ordering_tf_kernels.h5

vgg16-notop
https://github.com/fchollet/deep-learning-models/releases/download/v0.1/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5

vgg19
https://github.com/fchollet/deep-learning-models/releases/download/v0.1/vgg19_weights_tf_dim_ordering_tf_kernels.h5

vgg19-notop
https://github.com/fchollet/deep-learning-models/releases/download/v0.1/vgg19_weights_tf_dim_ordering_tf_kernels_notop.h5

resnet50
https://github.com/fchollet/deep-learning-models/releases/download/v0.2/resnet50_weights_tf_dim_ordering_tf_kernels.h5

resnet50-notop
https://github.com/fchollet/deep-learning-models/releases/download/v0.2/resnet50_weights_tf_dim_ordering_tf_kernels_notop.h5

xception
https://github.com/fchollet/deep-learning-models/releases/download/v0.4/xception_weights_tf_dim_ordering_tf_kernels.h5

xception-notop
https://github.com/fchollet/deep-learning-models/releases/download/v0.4/xception_weights_tf_dim_ordering_tf_kernels_notop.h5

mobilenet
https://github.com/fchollet/deep-learning-models/releases/download/v0.6/
后面加上mobilenet_%s_%d_tf.h5（包括顶层）或mobilenet_%s_%d_tf_no_top.h5（不包括顶层）
%s代表alpha，可取以下任一：1_0，7_5，5_0，2_5
%d代表图像大小，可取以下任一:128, 160, 192, 224

• 如alpha=1.0，图像为224，包括顶层。下载路径为https://github.com/fchollet/deep-learning-models/releases/download/v0.6/mobilenet_1_0_224_tf.h5
• 如alpha=0.50，图像为224，包括顶层。下载路径为https://github.com/fchollet/deep-learning-models/releases/download/v0.6/mobilenet_5_0_224_tf_no_top.h5

inception_v3
https://github.com/fchollet/deep-learning-models/releases/download/v0.5/inception_v3_weights_tf_dim_ordering_tf_kernels.h5

inception_v3-notop
https://github.com/fchollet/deep-learning-models/releases/download/v0.5/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5