staghado/vit.cpp

Vision Transformer (ViT) models for image classification converted to ggml format

Available models

Model	Disk	Mem	SHA
tiny	12 MB	~20 MB	25ce65ff60e08a1a5b486685b533d79718e74c0f
small	45 MB	~52 MB	7a9f85340bd1a3dcd4275f46d5ee1db66649700e
base	174 MB	~179 MB	a10d29628977fe27691edf55b7238f899b8c02eb
large	610 MB	~597 MB	5f27087930f21987050188f9dc9eea75ac607214

The models are pre-trained on ImageNet21k then finetuned on ImageNet1k with a patch size of 16 and an image size of 224.

For more information, visit:

https://github.com/staghado/vit.cpp