Dataset¶

Hub Datasets are dictionaries containing tensors. You can think of them as folders in the cloud. To store tensor in the cloud we first should put it in dataset and then store the dataset.

Store¶

To create and store dataset you would need to define tensors and specify the dataset dictionary.

from hub import dataset, tensor

tensor1 = tensor.from_zeros((20,512,512), dtype="uint8", dtag="image")
tensor2 = tensor.from_zeros((20), dtype="bool", dtag="label")

dataset.from_tensors({"name1": tensor1, "name2": tensor2})

dataset.store("username/namespace")

Load¶

To load a dataset from a central repository

from hub import dataset

ds = dataset.load("mnist/mnist")

Combine¶

You could combine datasets or concat them.

from hub import dataset

... 

#vertical
dataset.concat(ds1, ds2)

#horizontal
dataset.combine(ds1, ds2)

How to Upload a Dataset¶

For small datasets that would fit into your RAM you can directly upload by converting a numpy array into hub tensor. For complete example please check Uploading MNIST and Uploading CIFAR

For larger datasets you would need to define a dataset generator and apply the transformation iteratively. Please see an example below Uploading COCO. Please pay careful attention to meta(...) function where you describe each tensor properties. Please pay careful attention providing full meta description including shape, dtype, dtag, chunk_shape etc.

Dtag¶

For each tensor you would need to specify a dtag so that visualizer knows how draw it or transformations have context how to transform it.

Dtag	Shape	Types
default	any array	any
image	(width, height), (channel, width, height) or (width, height, channel)	int, float
text	used for label	str or object
box	[(4)]	int32
mask	(width, height)	bool
segmentation	(width, height), (channel, width, height) or (width, height, channel)	int
video	(sequence, width, height, channel) or (sequence, channel, width, height)	int, float
embedding
tabular
time
event
audio
pointcloud
landmark
polygon
mesh
document

Guidelines¶

Fork the github repo and create a folder under examples/dataset
Train a model using Pytorch

import hub
import pytorch

ds = hub.load("username/dataset")
ds = ds.to_pytorch()

# Implement a training loop for the dataset in pytorch
...

Train a model using Tensorflow

import hub
import tensorflow

ds = hub.load("username/dataset")
ds = ds.to_tensorflow()

# Implement a training loop for the dataset in tensorflow
...

Make sure visualization works perfectly at app.activeloop.ai

Final Checklist¶

So here is the checklist, the pull request.

Accessible using the sdk
Trainable on Tensorflow
Trainable on PyTorch
Visualizable at app.activeloop.ai
Pull Request merged into master

Issues¶

If you spot any trouble or have any question, please open a github issue.

API¶

class hub.dataset.Dataset(tensors: Dict[str, hub.collections.tensor.core.Tensor], metainfo={})¶

property citation¶: Dataset citation

property count¶: len of dataset (len of tensors across axis 0, yes, they all should be = to each other) Returns -1 if length is unknown

delete(tag, creds=None, session_creds=True) → bool¶: Deletes dataset given tag(filepath) and credentials (optional)

property description¶: Dataset description

property howtoload¶: Dataset howtoload

items()¶: Returns tensors

keys()¶: Returns names of tensors

property license¶: Dataset license

property meta¶: Dict of meta’s of each tensor meta of tensor contains all metadata for tensor storage

store(tag, creds=None, session_creds=True) → hub.collections.dataset.core.Dataset¶: Stores dataset by tag(filepath) given credentials (can be omitted)

to_pytorch(transform=None, max_text_len=30)¶

Transforms into pytorch dataset

Parameters

transform (func) – any transform that takes input a dictionary of a sample and returns transformed dictionary
max_text_len (integer) – the maximum length of text strings that would be stored. Strings longer than this would be snipped

to_tensorflow(max_text_len=30)¶

Transforms into tensorflow dataset

Parameters: max_text_len (integer) – the maximum length of text strings that would be stored. Strings longer than this would be snipped

values()¶: Returns tensors