resnet-protofiles-master
所属分类:人工智能/神经网络/深度学习
开发工具:Python
文件大小:34KB
下载次数:43
上传日期:2017-09-20 13:57:19
上 传 者:
半碗稀饭
说明: 卷积神经网络resnet18、resnet34、resnet50、resnet101、resnet152的配置文件(网络结构和解决方案文件)
(Convolution neural network resnet18, resnet34, resnet50, resnet101, resnet152 configuration file (network structure and solution file))
文件列表:
ResNet_101_deploy.prototxt (65439, 2017-04-15)
ResNet_101_train_val.prototxt (86517, 2017-04-15)
ResNet_152_deploy.prototxt (98034, 2017-04-15)
ResNet_152_train_val.prototxt (128819, 2017-04-15)
ResNet_18_train_val.prototxt (18440, 2017-04-15)
ResNet_34_train_val.prototxt (31670, 2017-04-15)
ResNet_50_deploy.prototxt (38518, 2017-04-15)
ResNet_50_train_val.prototxt (43888, 2017-04-15)
generator.py (11011, 2017-04-15)
solver.prototxt (317, 2017-04-15)
Current config for 8 GPUs with 32 mini-batch size each.
The layer names are designed to match MSRA released pre-trained models to allow for finetuning. You may need to enable/disable bias on `conv1` to use these prototxts with some pretrained models.
Running the ResNet-50 as-is gets a few percent lower accuracy than MSRA if done without random reshape & crop.
近期下载者:
相关文件:
收藏者: