metadata
library_name: keras-hub
This is a ViT model uploaded using the KerasHub library and can be used with JAX, TensorFlow, and PyTorch backends.
Model config:
- name: vi_t_backbone
- trainable: True
- image_shape: [224, 224, 3]
- patch_size: 14
- num_layers: 32
- num_heads: 16
- hidden_dim: 1280
- mlp_dim: 5120
- dropout_rate: 0.0
- attention_dropout: 0.0
- layer_norm_epsilon: 1e-06
- use_mha_bias: True
- use_mlp_bias: True
This model card has been generated automatically and should be completed by the model author. See Model Cards documentation for more information.