metadata

license: apache-2.0
tags:
  - mlx
  - mlx-image
  - vision
  - image-classification
datasets:
  - imagenet-1k
library_name: mlx-image

vit_base_patch16_384.swag_e2e

A Vision Transformer image classification model. Weights are learned with SWAG on ImageNet-1k data.

Disclaimer: This is a porting of the torchvision model weights to Apple MLX Framework.

How to use

pip install mlx-image

Here is how to use this model for image classification:

from mlxim.model import create_model
from mlxim.io import read_rgb
from mlxim.transform import ImageNetTransform

transform = ImageNetTransform(train=False, img_size=384)
x = transform(read_rgb("cat.png"))
x = mx.expand_dims(x, 0)

model = create_model("vit_base_patch16_384.swag_e2e")
model.eval()

logits = model(x)

You can also use the embeds from layer before head:

from mlxim.model import create_model
from mlxim.io import read_rgb
from mlxim.transform import ImageNetTransform

transform = ImageNetTransform(train=False, img_size=384)
x = transform(read_rgb("cat.png"))
x = mx.expand_dims(x, 0)

# first option
model = create_model("vit_base_patch16_384.swag_e2e", num_classes=0)
model.eval()

embeds = model(x)

# second option
model = create_model("vit_base_patch16_384.swag_e2e")
model.eval()

embeds = model.features(x)

Model Comparison

Explore the metrics of this model in mlx-image model results.