kevinlin311tw/caffe-cvprw15

How I can train myself datasets with ImageNe pretrained model ?

aiaini66 opened this issue · 4 comments

I have read the README.md file and see your words:"You need to convert the dataset into leveldb format using "create_imagenet.sh". We will show you how to do this. To be continued.". Now I want to train myself Datasets with ImageNe pretrained model ,if there are any instruction about how to convert myself Datasets format? I'm forward to your reply!

  1. Change the paths in create_imagenet.sh with your dataset path.
    $ vi /examples/cvprw15-cifar10/create_imagenet.sh

  2. Run script prepare_train.sh to download ImageNet pretrained model and convert myself dataset to leveldb format.
    $ ./prepare_train.sh

  3. Change the source with your leveldb source in training script.
    $ vi /examples/cvprw15-cifar10/train_48.sh

  4. Go to the folder /examples/cvprw15-cifar10, and run the training script:

$ cd /examples/cvprw15-cifar10
$ chmod 777 train_48.sh
$ ./train_48.sh

  1. You will finally get your model named KevinNet_CIFAR10_48_iter_xxxxxx.caffemodel. To use your model, modify the model_file in demo.m to link to your model:

    model_file = './YOUR/MODEL/PATH/filename.caffemodel';

  2. Launch matlab and run demo.m.

thanks a lot! If I have any questions in the further work ,I'll ask you for help ! best wish!

uumom commented

The baidu link of model and image resource is invalid, can anyone share it?

About the baidu link, I found that my baidu account is dead since Baidu request new authentication. It only accept China's phone number, but I don't have one. Sorry for the convenience.