SmartCacheDatset does not see the whole train data when cache_rate is less than 1

[YazdanSalimi]

Thank you for the great platform you developed.
I am reproducing the spleen 3d segmentation example from monai tutorials.
The only thing I change is the train dataset from CacheDataset to SmartChachDataset.
the original code is : train_ds = CacheDataset(data=train_files, transform=train_transforms, cache_rate=1.0, num_workers=4)
The modified code is: train_ds = SmartCacheDataset(data=train_files, transform=train_transforms, cache_rate=.5, num_workers=0)

after this change, if the chach_rate is less than one (e.g., 0.1) the training loop only sees .1 of the cases. this problem does not exist with CacheDataset. Because this is the exact code of one of the tutorials, I did not copy my code here.
Thank you

[KumoLiu]

Hi @YazdanGit, thanks for your interest here.
That, I believe, is the purpose of this dataset.

MONAI/monai/data/dataset.py

Lines 936 to 957 in 6f5cea3

	At any time, the cache pool only keeps a subset of the whole dataset. In each epoch, only the items
	in the cache are used for training. This ensures that data needed for training is readily available,
	keeping GPU resources busy. Note that cached items may still have to go through a non-deterministic
	transform sequence before being fed to GPU. At the same time, another thread is preparing replacement
	items by applying the transform sequence to items not in cache. Once one epoch is completed, Smart
	Cache replaces the same number of items with replacement items.
	Smart Cache uses a simple `running window` algorithm to determine the cache content and replacement items.
	Let N be the configured number of objects in cache; and R be the number of replacement objects (R = ceil(N * r),
	where r is the configured replace rate).
	For more details, please refer to:
	https://docs.nvidia.com/clara/clara-train-archive/3.1/nvmidl/additional_features/smart_cache.html
	If passing slicing indices, will return a PyTorch Subset, for example: `data: Subset = dataset[1:4]`,
	for more details, please check: https://pytorch.org/docs/stable/data.html#torch.utils.data.Subset
	For example, if we have 5 images: `[image1, image2, image3, image4, image5]`, and `cache_num=4`, `replace_rate=0.25`.
	so the actual training images cached and replaced for every epoch are as below::
	epoch 1: [image1, image2, image3, image4]
	epoch 2: [image2, image3, image4, image5]
	epoch 3: [image3, image4, image5, image1]
	epoch 3: [image4, image5, image1, image2]
	epoch N: [image[N % 5] ...]

(Convert to discussion for now, feel free to create another one if you still have some issues.)

[YazdanSalimi]

Thank you for the response.
What I observed by SmartChachDataset was loading [omage1, image2, image3, image4] for every epoch. I tested it by printing the image["image_meta_dict"]["file_name_or_obj"] every time and it only sees the images the same as the 1st epoch.
Thank you.