Use external zip files

Clean old url pointers to split and labels data

Comment task_template in dataset info

Change metadata urls from dummy -> data files

Upload split and label info

Change labels to just multilabels

Reuse a line

Test multilabel

Clean NIH-Chest-X-ray-dataset.py script

Fix pandas subset series

Read csv file and get the label for each image id

Add label file and import read_csv from pandas

Comment multi-label

Check results

check label type

Modify label using dataset.features.Sequence() -> multi-label classification

Check multi-index list encode for multilabels

Fix typo: daatasets -> datasets

Check multi-labels

Add a hardcode valid class label (_NAMES)

Fix typo: file_namae -> file_name

Add the _generate_examples method

Fix typo

Add condition to check if an image path belong to train-val-set

Iterate for urls in each zip

fix typo

Add logger

return dataset lists as iter(dataset_lists)

fix typo

Fix tab typo

Fix tab format

Fix typo NIH

Fix typo

iterate trought image names and store it in a list for each dataset

try csv parse

Create dummy data for testing

Fix labels names in dataset features

Change name to the data loader script

init dataset loader script - info

initial commit