OPEN
QEV Compact decision model
QEV Compact prunes two multimodal models to 466M and 1.31B parameters. Read its open weights, public vision results and calibration limits.
By Ken Jo · Hugging Face weight release: 9 Nov 2026
Two pruned QEV v3 models for text and image questions. The maker removes eight of 24 language-model blocks and trims the input vocabulary to 32,768 rows, then distils a decision readout. They return option probabilities, a typed answer and a separate abstain flag. Ken Jo's QEV family is separate from labqoat's Qev 0.8B.
Quick answers
What is QEV Compact by Ken Jo?
QEV Compact: Two pruned QEV v3 models for text and image questions.
Who makes QEV Compact?
Facts
- Sizes
- 466,313,408 and 1,314,213,824 parameters, per the cards
- Inputs
- English text and up to four images; Choice, Noul and Score
- Context
- 8,192 input tokens; over-budget requests refused
- Runtime
- Bundled backbone and head; pinned qev Python loader on CUDA, accurate mode only
- Licence
- Apache-2.0 files included; training datasets retain their own terms
Claims and sources
- The maker reports public vision index scores of 52.52 and 62.15 on 12,210 rows across 11 benchmarks, versus 52.70 and 62.27 for the corresponding full-depth parents. These are public-only normalized scores, not accuracy percentages; private-half results are pending. vendor reportedsource
- On an RTX 4060 Ti 8 GB at batch 1, the maker reports peak GPU memory of 1.15 GiB and 2.76 GiB, with median p50 timings of 54.4 ms and 99.0 ms across 47 request profiles. Calibration and abstention thresholds were fitted on development data; the empirical error target is not a guarantee on new data. vendor reportedsource
Links
Source: 2B compact card and weights → Source: 0.8B compact card and weights → Source: Licence →
Related projects
No related projects are listed in the registry.