--- task_categories: - image-classification --- # AutoTrain Dataset for project: image-attribute-prediction ## Dataset Description This dataset has been automatically processed by AutoTrain for project image-attribute-prediction. ### Languages The BCP-47 code for the dataset's language is unk. ## Dataset Structure ### Data Instances A sample from this dataset looks as follows: ```json [ { "image": "<261x300 RGB PIL image>", "target": 0 }, { "image": "<300x300 RGB PIL image>", "target": 0 } ] ``` ### Dataset Fields The dataset has the following fields (also called "features"): ```json { "image": "Image(decode=True, id=None)", "target": "ClassLabel(names=['Paisley_Floral_A-Line_Dress', 'Paisley_Maxi_Cami_Dress'], id=None)" } ``` ### Dataset Splits This dataset is split into a train and validation split. The split sizes are as follow: | Split name | Num samples | | ------------ | ------------------- | | train | 59 | | valid | 16 |