A compact vision-language model for document and screenshot understanding.
Q4_K_M default
tokens
Canonical metadata is separated from immutable version and Run evidence.
Lyra Vision 6B demonstrates multimodal metadata, image-aware tasks and runtime-specific compatibility.
1 published Runs; each result links to its exact conditions and evidence level.
Benchmark claims remain tied to immutable revisions even after the repository changes.
Public preview with document-layout improvements.
