MM1
MM1 is Apple Research’s multimodal LLM blueprint: a vision encoder feeding a text decoder via cross-attention, pretrained on a balanced mix of image–caption, interleaved image–text, and text-only data. It highlights how data quality, interleaving, and resolution—not just scale—drive strong OCR, document/chart reasoning, and grounded visual answers.
- Type
- Text
- Weights
- Open
Model hub
Ashish9879/mm1
- Repo
- Ashish9879/mm1
- Access
- Public
More from Apple
9
-
198k DL
DFN2B-CLIP-ViT-B-16
License apple-amlr · 198,144 downloads · by apple
-
51k DL
DFN5B-CLIP-ViT-H-14-378
License apple-amlr · 509,768 downloads · by apple
-
FastVLM
Text
FastVLM is Apple’s lightweight vision-language model built for real-time multimodal apps. It ingests images a…
-
Image Playground
Image
Image Playground is Apple’s on-device generative image tool inside Apple Intelligence. It creates playful vis…
-
LiTo
Multimodal
LiTo is Apple’s research image-to-3D model and tokenizer for generating 3D objects with unified geometry and…
-
Manzano
Multimodal
Manzano is Apple’s unified multimodal model that shares a hybrid vision tokenizer for both image understandin…
-
MM1.5
Text
MM1.5 is Apple Research’s refinement of the MM1 multimodal recipe. It keeps the same encoder–decoder architec…
-
1.2M DL
OpenELM-1_1B-Instruct
Text
text generation · License apple-amlr · 1,204,778 downloads · by apple
About Apple
1