TIGER-Lab/Mantis-8B-Idefics2
Image-Text-to-Text
•
Updated
•
655
•
13
Mantis
Mantis model family optimized for multi-image reasoning with interleaved text/image format
Note Current SoTA Mantis variant
-
TIGER-Lab/Mantis-8B-clip-llama3
Image-Text-to-Text • Updated • 334 • 1
TIGER-Lab/Mantis-8B-siglip-llama3
Image-Text-to-Text • Updated • 16.7k • 32Note Current SoTA Mantis variant without multi-image pre-training
TIGER-Lab/Mantis-Instruct
Viewer • Updated • 999k • 2.53k • 32Note Our training dataset
TIGER-Lab/Mantis-Eval
Viewer • Updated • 217 • 368 • 6Note Curated evaluation benchmark for multi-image scenarios
-
MANTIS: Interleaved Multi-Image Instruction Tuning
Paper • 2405.01483 • Published • 6 -
Running on Zero25👁
Mantis
Multimodal Language Model
-
TIGER-Lab/Mantis-llava-7b
Image-Text-to-Text • Updated • 10 • 15 -
TIGER-Lab/Mantis-bakllava-7b
Image-Text-to-Text • Updated • 31 • 5 -
TIGER-Lab/Mantis-8B-Fuyu
Text Generation • Updated • 64 • 4 -
TIGER-Lab/OBELICS-100K
Viewer • Updated • 212k • 45 • 2