ICGen

Installation

pip install icgen

for a development installation see CONTRIBUTING.md

Usage

Sampling Datasets

import icgen
dataset_generator = icgen.ICDatasetGenerator(
  data_path="datasets",
  min_resolution=16,
  max_resolution=512,
  max_log_res_deviation=1,  # Sample only 1 log resolution from the native one
  min_classes=2,
  max_classes=100,
  min_examples_per_class=20,
  max_examples_per_class=100_000,
)
dev_data, test_data, dataset_info = dataset_generator.get_dataset(
    dataset="cifar10", augment=True, download=True
)

The augment parameter controls whether the original dataset is modified.

Options only affect sampling with augment=True and the min max ranges do not filter datasets.

The data is left at the original resolution, so it can be resized under user control. This is necessary to for example avoid resizing twice which can hurt performance.

You can also sample from a list of datasets

dataset_generator.get_dataset(datasets=["cifar100", "emnist/balanced"], download=True)

We provide some lists of available datasets

import icgen
icgen.DATASETS_TRAIN
icgen.DATASETS_VAL
icgen.DATASETS_TEST
icgen.DATASETS

or on the commandline you can get the names with

python -m icgen.dataset_names

Downloading Datasets Before Execution

To download datasets ahead of time you can run

python -m icgen.download --data_path DATA_PATH --datasets D1 D2 D3

or directly download a complete group

python -m icgen.download --data_path DATA_PATH --dataset_group GROUP  # all, train, dev, test

Alternatively, you can also use the download=True flag of the dataset_generator.get_dataset function.

Reconstructing and Distributing Tasks

In distributed applications it may be necessary to sample datasets on one machine and then use them on another one. Conversely, for reproducibility it may be necessary to store the exact dataset which was used. For these cases icgen uses a dataset identifier which uniquely identifies datasets.

License

MIT