Download SOURCE.json from AndeyTait/JevForge-Mind2Web: direct link, hf CLI and curl.
- Browser
- Download file 685 Bytes
-
https://huggingface.co/datasets/AndeyTait/JevForge-Mind2Web/resolve/main/SOURCE.json
- Command line
-
hf download hf://datasets/AndeyTait/JevForge-Mind2Web/SOURCE.json
-
curl -L -o SOURCE.json https://huggingface.co/datasets/AndeyTait/JevForge-Mind2Web/resolve/main/SOURCE.json
685 Bytes
| { | |
| "dataset": "LangAGI-Lab/Mind2Web-axtree-cleaned-lite", | |
| "dataset_revision": "080a4b67d0746270166a57f2a4096a6b4129523c", | |
| "upstream_project": "OSU-NLP-Group/Mind2Web", | |
| "upstream_license": "CC-BY-4.0", | |
| "source_split": "train", | |
| "official_test_archive_included": false, | |
| "conversion": { | |
| "schema_version": "openjev-record-v1", | |
| "seed": 17, | |
| "split_unit": "website", | |
| "max_candidates": 12, | |
| "state_chars": 2600, | |
| "description_chars": 88 | |
| }, | |
| "split_semantics": { | |
| "test": "JevForge internal held-out websites; not an official Mind2Web test split", | |
| "ood": "JevForge internal held-out-site challenge; not the official Mind2Web test-domain split" | |
| } | |
| } | |