sanps
/

fVLM-135M

Image-Text-to-Text

vision-language

video-understanding

foveated-attention

Model card Files Files and versions

Commit History

Feb 21

18:38
Update model card: drop POPE, correct ScienceQA to 36.0%, add inference modes table

sanps

Verified signature
05571a0
17:44
docs: update benchmarks — add SmolVLM2 comparison, remove Val10K, split video/image

sanps

Verified signature
6164fab
17:22
docs: add benchmark results table + SmolVLM2 comparison

sanps

Verified signature
27c3839

Feb 20

06:25
Upload fVLM-135M: Foveated Vision-Language Model (Stage 3 DPO)

sanps

Verified signature
6d320d6
06:25
initial commit

sanps

Verified signature
561dc59