origami-ml/jsynth-data

--- configs: - config_name: adult data_files: - split: train path: adult/train.jsonl - split: test path: adult/test.jsonl - config_name: diabetes data_files: - split: train path: diabetes/train.jsonl - split: test path: diabetes/test.jsonl - config_name: electric_vehicles data_files: - split: train path: electric_vehicles/train.jsonl - split: test path: electric_vehicles/test.jsonl - config_name: ddxplus data_files: - split: train path: ddxplus/train.jsonl - split: test path: ddxplus/test.jsonl --- # jsynth-data Preprocessed datasets for Origami tabular/JSON synthesis experiments. Each configuration is an independent dataset with its own schema and train/test split. ## Datasets | Dataset | Train | Test | Type | |---------|-------|------|------| | adult | 32,561 | 16,281 | Tabular | | diabetes | 61,059 | 20,354 | Tabular | | electric_vehicles | 189,010 | 21,001 | Tabular | | ddxplus | 1,025,602 | 134,529 | Semi-structured | ## Usage ```python from huggingface_hub import hf_hub_download path = hf_hub_download( repo_id="origami-ml/jsynth-data", filename="adult/train.jsonl", repo_type="dataset", ) ``` Or with the `datasets` library: ```python from datasets import load_dataset ds = load_dataset("origami-ml/jsynth-data", "adult") ```