MCPcopy Create free account
hub / github.com/alinlab/SelfPatch / vit_tiny

Function vit_tiny

selfpatch_vision_transformer.py:344–348  ·  view source on GitHub ↗
(patch_size=16, **kwargs)

Source from the content-addressed store, hash-verified

342 return output
343
344def vit_tiny(patch_size=16, **kwargs):
345 model = VisionTransformer(
346 patch_size=patch_size, embed_dim=192, depth=12, num_heads=3, mlp_ratio=4,
347 qkv_bias=True, norm_layer=partial(nn.LayerNorm, eps=1e-6), **kwargs)
348 return model
349
350def vit_small(patch_size=16, **kwargs):
351 model = VisionTransformer(

Callers

nothing calls this directly

Calls 1

VisionTransformerClass · 0.85

Tested by

no test coverage detected