OpenMOSS-Team
/

MOSS-VL-Realtime-FP8

Video-Text-to-Text

feature-extraction

video-understanding

compressed-tensors

Model card Files Files and versions

MOSS-VL-Realtime-FP8 / processing_moss_vl.py

Commit History

per-sample media budgets in processor; per-sample MRoPE position ids for batched vision inputs

e0213fe
verified

CCCCyx commited on 12 days ago

fix: apply query RoPE when reusing cached vision KV; align processor with the 0708 standard

5cf52b6
verified

CCCCyx commited on 24 days ago

Add files using upload-large-folder tool

d7c049e
verified

CCCCyx commited on Aug 11