License: nvidia-evaluation-data-license
May 19 21:13 viewer: add configs YAML + metadata.jsonl for 2dbbox & dense_captioning Enables one HF dataset-viewer tab per task. Adds a `configs:` block to README.md and two metadata.jsonl files so the image / video tasks are browsable. CHANGELOG entry recorded under 2026-05-19. 00:24 Dataset card refresh + promote CHANGELOG.md as primary release history (#20 ) - Dataset card refresh + promote CHANGELOG.md as primary release history (d23be381d2393df107f42da1b87961a45c43f981) 00:05 README: replace dataset_info with per-task configs so HF viewer loads each schema independently (#21 ) - README: replace dataset_info with per-task configs so HF viewer loads each schema independently (4ead2a2a6c86bab090a7ffc043c9cb8b9c3b1319) May 18 22:41 VANTAGE-Bench update: prompts + question-only annotations + Vantage2DPointing rename + test split + data/README + CHANGELOG (#19 ) vpraveen-nv +1: arihantj - VANTAGE-Bench update: prompts + question-only annotations + Vantage2DPointing rename + test split + data/README + CHANGELOG (4e1fb24231ecca2e1972186b376ac60e0139977d)
- Revert README.md to main (drop YAML configs change from PR) (89ef7072457bd9fe0f50625f128e9950d70ea769)
- Unify annotation schemas (EV bcq, VQA + Temporal field union) so HF viewer can load (3042ce1041c007c515f2b10b7fc9602bf90d17e3)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> 21:57 Declare test-only split in dataset_info frontmatter May 12 00:33 Point License section to LICENSE.md in dataset 00:32 Remove paper citation TBD reminder 00:27 Rename LICENSE -> LICENSE.md so it resolves from frontmatter May 8 22:40 croissant: ensure subset has the latest version with license/citeAs/version/datePublished/RAI (#17 ) - croissant: ensure subset has the latest version with license/citeAs/version/datePublished/RAI (ac8ebc37b9aa48f568060e806877f5e683143aae) 07:26 Rename LICENSE.md -> LICENSE (#16 ) - Rename LICENSE.md -> LICENSE (a3cdecde37ab5ff9c8ad4730ae3b72fa9e75013e) 04:09 croissant: keep only rai:-prefixed RAI fields (deduplicate) (#14 ) vpraveen-nv +1: arihantj - croissant: keep only rai:-prefixed RAI fields (deduplicate) (1fe4fcf2334feb1b9944bafaa1f43d4290b9ba8c)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> 03:11 croissant.json: cumulative update (RAI dual keys, citation, new title/authors) (#12 ) vpraveen-nv +1: arihantj - croissant.json: cumulative update (RAI dual keys, citation, new title/authors) (e2c0da9b2a3a9f06e9ac4a2f6afdf7710d7695ba)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> 01:31 croissant: dual rai: prefixed + unprefixed RAI keys for validator compatibility (#9 ) vpraveen-nv +1: arihantj - croissant: dual rai: prefixed + unprefixed RAI keys for validator compatibility (d0518361c390a161f664ec2da8a0740309dc6203)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> 00:31 croissant.json: unprefix RAI fields, add @context aliases (NeurIPS RAI compatibility) (#8 ) vpraveen-nv +1: arihantj - croissant.json: unprefix RAI fields, add @context aliases (NeurIPS RAI compatibility) (ed354f2d37fc01bd99411c59b024200613469746)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> 00:31 croissant.json: split aggregated recordSets to one per fileset (source cardinality fix) (#7 ) vpraveen-nv +1: arihantj - croissant.json: split aggregated recordSets to one per fileset (source cardinality fix) (7f9d9a7a1c8aeeb43e8ef885d3187bff50b58a4f)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> May 7 23:31 Update Croissant 1.1 metadata: fix forbidden chars, paths, recordSet IDs (#6 ) vpraveen-nv +1: arihantj - Update Croissant 1.1 metadata: fix forbidden chars, paths, recordSet IDs (caf1617f384db5469f8099e5422ef98669eb392b)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> 06:50 Add Croissant 1.1 metadata (croissant.json) (#4 ) vpraveen-nv +1: arihantj - Add Croissant 1.1 metadata (croissant.json) (5fb96790809bc050d5b1ec4215d1a35368867080)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> 06:49 Restructure data/: drop Spatial, rename to dense_captioning / event_verification / pointing / referring / temporal_localization / vqa (#5 ) vpraveen-nv +1: arihantj - Restructure data/: drop Spatial, rename to dense_captioning / event_verification / pointing / referring / temporal_localization / vqa (ef2b4b0a9759fe7ec55967477eae3648413c70a5)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> 05:32 Replace data/ with new 4 GB proportional subset (#3 ) vpraveen-nv +1: arihantj - Replace data/ with new 4 GB proportional subset (513038ef563997304c4e3398758b7b4cde720e61)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> May 6 18:30 Remove annotations/ — subset hosts media only (#2 ) vpraveen-nv +1: arihantj - Remove annotations/ — subset hosts media only (0812fc0695c509e639b5876d6f11f310a0f9d4c2)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> May 5 23:40 Move task overview figure from Dataset Description to References 23:10 Mirror VANTAGE-Bench dataset card and add markdown LICENSE 07:37 Add VANTAGE-Bench subset (media + annotations + MANIFEST) (#1 ) vpraveen-nv +1: arihantj - Add VANTAGE-Bench subset (media + annotations + MANIFEST) (79db94d24ccab8e3d133bf6245a9aed98e66f17d)
Co-authored-by: Arihant Jain <arihantj@users.noreply.huggingface.co> 06:44 initial commit 
