OpenMOSS-Team
/

MOSS-VL-Base-0708

Video-Text-to-Text

feature-extraction

Video-Understanding

Image-Understanding

vision-language

Model card Files Files and versions

MOSS-VL-Base-0708

Commit History

Sep 23 30 days ago

03:54
per-sample media budgets in processor; per-sample MRoPE position ids for batched vision inputs

CCCCyx

Verified signature
4f1c992
03:54
per-sample media budgets in processor; per-sample MRoPE position ids for batched vision inputs

CCCCyx

Verified signature
f337b2e

Sep 22

09:17
sync offline system prompts with training data: no_thinking -> no system message, deep_thinking -> <think>/<answer> tags

CCCCyx

Verified signature
f5137d7

Sep 10

03:22
fix: apply query RoPE in cross-attention when reusing cached vision KV during decode

CCCCyx

Verified signature
417395c

Sep 7

10:48
Unify MOSS-VL processor and inference preprocessing defaults⁠

CCCCyx

Verified signature
ba1bee7

Aug 18

07:28
Upload README.md with huggingface_hub

CCCCyx

Verified signature
e737f19

Aug 10

04:26
Update video_processing_moss_vl.py

CCCCyx

Verified signature
da8796b

Aug 4

02:33
Update README.md

CCCCyx

Verified signature
1f4da11

Jul 15

02:50
Update README.md

CCCCyx

Verified signature
d3dae3c

Jul 14

09:29
Upload benchmark-offline.png

CCCCyx

Verified signature
deccac3
09:28
Delete assets/benchmark-offline.png

CCCCyx

Verified signature
b9789d8
08:54
Upload benchmark-offline.png

CCCCyx

Verified signature
77f9a20
08:54
Delete assets/benchmark-offline.png

CCCCyx

Verified signature
3068617
03:03
Update model card and assets

CCCCyx

Verified signature
9dc4e66

Jul 13

13:16
Upload folder using huggingface_hub

CCCCyx

Verified signature
a2ae4e5
11:36
initial commit

CCCCyx

Verified signature
0cb14a4