← MIC coverage of Transformers model types

apple/mobilevit-xx-small

huggingface-mic/mobilevit-xx-small-mic@857be9b7c6ea ← apple/mobilevit-xx-small@6703997f9e94 · MobileViTForImageClassification · real weights

CI: passing now; 11 failed checks in 30 days · MIC failures dashboard

Original1.3Mparameters
Fixture1.3Mparameters
Layer stacks1compared
Findings0to review

✓ Same architecture, smaller

Every module class, layer kind and checked property of the original is in the fixture.

Layer stacks

One square per layer, colored by kind: the layer's module classes and type attributes. Hover a square for its index and kind.

mobilevit.encoder.layer

Original · 5
Fixture · 5
KindOriginalFixture
MobileNetLayer22✓ kept
LayerNorm, Linear, Attention …33✓ kept

Where the parameters are

Share of each model's parameters per component. Vocabulary and patch sizes are kept, so embeddings usually take a larger share of the fixture.

Original
Fixture
ComponentOriginalShareFixtureShare
mobilevit.encoder.layer924.3K72.7%924.3K72.7%
classifier321.0K25.2%321.0K25.2%
mobilevit.conv_1x1_exp26.2K2.1%26.2K2.1%
mobilevit.conv_stem4640.0%4640.0%

Preserved properties

ConfigPropertyOriginalFixture
modelmodel_typemobilevitmobilevit✓ kept
modelhidden_actsilusilu✓ kept
modelimage_size256256✓ kept
modelpatch_size22✓ kept
modelnum_channels33✓ kept

Module classes

Instances of each class in the built model (prefix MobileViT omitted).

ClassOriginalFixture
Linear5555
Conv2d3535
ConvLayer3535
SiLUActivation3434
BatchNorm2d3232
Dropout2828
LayerNorm2121
Attention99
Intermediate99
Output99
SelfAttention99
SelfOutput99
TransformerLayer99
InvertedResidual77
ModuleList66
Layer33
Transformer33
MobileNetLayer22
Encoder11
ForImageClassification11
Model11

Config changes

0 fields changed, 30 unchanged.

FieldOriginalFixture