forked from huggingface/datasets
-
Notifications
You must be signed in to change notification settings - Fork 0
/
dataset_infos.json
1 lines (1 loc) 路 18.1 KB
/
dataset_infos.json
1
{"simple": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "simple", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 801481, "num_examples": 4182, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 3224035, "num_examples": 16728, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 4025516, "size_in_bytes": 22706428}, "addprim_jump": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "addprim_jump", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 1511337, "num_examples": 7706, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 2541119, "num_examples": 14670, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 4052456, "size_in_bytes": 22733368}, "addprim_turn_left": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "addprim_turn_left", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 170516, "num_examples": 1208, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 3917086, "num_examples": 21890, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 4087602, "size_in_bytes": 22768514}, "filler_num0": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "filler_num0", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 330528, "num_examples": 1173, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 2518738, "num_examples": 15225, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 2849266, "size_in_bytes": 21530178}, "filler_num1": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "filler_num1", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 330528, "num_examples": 1173, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 2808968, "num_examples": 16290, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 3139496, "size_in_bytes": 21820408}, "filler_num2": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "filler_num2", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 330528, "num_examples": 1173, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 3112734, "num_examples": 17391, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 3443262, "size_in_bytes": 22124174}, "filler_num3": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "filler_num3", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 330528, "num_examples": 1173, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 3419644, "num_examples": 18528, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 3750172, "size_in_bytes": 22431084}, "length": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "length", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 1346688, "num_examples": 3920, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 2678828, "num_examples": 16990, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 4025516, "size_in_bytes": 22706428}, "template_around_right": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "template_around_right", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 1231437, "num_examples": 4476, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 2518738, "num_examples": 15225, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 3750175, "size_in_bytes": 22431087}, "template_jump_around_right": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "template_jump_around_right", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 330528, "num_examples": 1173, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 3419644, "num_examples": 18528, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 3750172, "size_in_bytes": 22431084}, "template_opposite_right": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "template_opposite_right", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 859623, "num_examples": 4476, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 2950102, "num_examples": 15225, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 3809725, "size_in_bytes": 22490637}, "template_right": {"description": "SCAN tasks with various splits.\n\nSCAN is a set of simple language-driven navigation tasks for studying\ncompositional learning and zero-shot generalization.\n\nSee https://github.com/brendenlake/SCAN for a description of the splits.\n\nExample usage:\ndata = datasets.load_dataset('scan/length')\n", "citation": "\n@inproceedings{Lake2018GeneralizationWS,\n title={Generalization without Systematicity: On the Compositional Skills of\n Sequence-to-Sequence Recurrent Networks},\n author={Brenden M. Lake and Marco Baroni},\n booktitle={ICML},\n year={2018},\n url={https://arxiv.org/pdf/1711.00350.pdf},\n}\n", "homepage": "https://github.com/brendenlake/SCAN", "license": "", "features": {"commands": {"dtype": "string", "id": null, "_type": "Value"}, "actions": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "scan", "config_name": "template_right", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 718083, "num_examples": 4476, "dataset_name": "scan"}, "train": {"name": "train", "num_bytes": 3133327, "num_examples": 15225, "dataset_name": "scan"}}, "download_checksums": {"https://github.com/brendenlake/SCAN/archive/master.zip": {"num_bytes": 18680912, "checksum": "ae9038488f880aedf9a5ed9e9693f22f0a3e33d6a4e3aa4d1e8006d9c038594a"}}, "download_size": 18680912, "dataset_size": 3851410, "size_in_bytes": 22532322}}