Sign inGet started
Dataset · Hugging Face

Kkuntal990/LIVR_mixed

LIVR_mixed (v2) Curated multimodal data for training a Qwen2.5-VL self-reflection RL pipeline, plus the held-out LIVR splits and three external benchmarks (BLINK + PixMo-Count + VSP) used to evaluate it. Layout train/ # LIVR train (9 tasks × 1000) metadata.jsonl livr_v2_manifest.json images/<task>/... # ~8.7 GB livr_eval/ # LIVR's own held-out val + test (8 tasks; counting → pixmo_count_eval)… See the full description on the dataset page: https://huggingface.co/datasets/Kkuntal990/LIVR_mixed.

Author / organizationKkuntal990
License (as declared)cc-by-4.0
Downloads419 (as reported by Hugging Face)
Likes0
Created21 April 2026
Last updated at source20 May 2026
visual-reasoningmultimodalvqamcqcountingspatial-planninggrpoself-reflection
Embed this entry
<iframe src="https://silytics.si/embed/item/53c0cdec-d6cc-402d-b3ff-50a9949c7869" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 3 days

419 now (+3 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

Recent changes

  • 16 Oct 2026 Description: LIVR_mixed (v2) Curated multimodal data for training a Qwen2.5-VL self-reflection RL pipeline, plus the held-out LIVR splits and three external benchmarks (BLINK + PixMo-Count + VSP) used to evaluate it. Layout train/ # LIVR train (9 tasks → LIVR_mixed (v2) Curated multimodal data for training a Qwen2.5-VL self-reflection RL pipeline, plus the held-out LIVR splits and three external benchmarks (BLINK + PixMo-Count + VSP) used to evaluate it. Layout train/ # LIVR train (9 tasks

Changes detected by Silytics between refreshes of the source.

More from Kkuntal990

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.