about the datasets and dataloader

[liujianzhao6328057]

Thanks for your job. I began to enter the “few shot incremental” filed and not familiar with the datasets. Could you please release the datasets.py and dataloader first for reference? Thank you very much.

[xyutao]

An example dataloader/dataloader.py has been added. Noting that the original code is written with mxnet.

[liujianzhao6328057]

Thank you. The dataloader is the same as that of few shot dataloader. Is the difference between the few shot and few shot incremental the way of training? the dataset yields one batch/task (N way K shot), and dataloader with batch size of M yileds a task sequence containing M tasks. Then we need to train the M batches/tasks sequentially and test the performance of all the M tasks at the end of the task sequence. Is my understanding right?

[xyutao]

After learning M tasks, the model is tested on all the test images of the N * M classes, which is N * M * 100 (per class) for CIFAR100

[liujianzhao6328057]

Thanks