ViTAE-B pretrained model does not matched with ViTDet-ViTAE-Base-100e.py?

[foolhard]

Hello,

I use your ViTAE-B pretrained model ("convert_base_22k.pth"）to train ViTDet-ViTAE.

But found the model is not matched.

Which is the correct pretrained ViTAE-B model?

Below is the log:

`unexpected key in source state_dict: layers.0.RC.cls_token, layers.0.RC.pos_embed, layers.0.RC.PRM.proj.weight, layers.0.RC.PRM.proj.bias, layers.2.NC.0.norm1.weight, layers.2.NC.0.norm1.bias, layers.2.NC.0.attn.qkv.bias, layers.2.NC.0.attn.qkv.weight, layers.2.NC.0.attn.proj.weight, layers.2.NC.0.attn.proj.bias, layers.2.NC.0.norm2.weight, layers.2.NC.0.norm2.bias, layers.2.NC.0.mlp.fc1.weight, layers.2.NC.0.mlp.fc1.bias, layers.2.NC.0.mlp.fc2.weight, layers.2.NC.0.mlp.fc2.bias, layers.2.NC.0.PCM.0.weight, layers.2.NC.0.PCM.0.bias, layers.2.NC.0.PCM.1.weight, layers.2.NC.0.PCM.1.bias, layers.2.NC.0.PCM.1.running_mean, layers.2.NC.0.PCM.1.running_var, layers.2.NC.0.PCM.1.num_batches_tracked, layers.2.NC.0.PCM.3.weight, layers.2.NC.0.PCM.3.bias, layers.2.NC.1.norm1.weight, layers.2.NC.1.norm1.bias, layers.2.NC.1.attn.qkv.bias, layers.2.NC.1.attn.qkv.weight, layers.2.NC.1.attn.proj.weight, layers.2.NC.1.attn.proj.bias, layers.2.NC.1.norm2.weight, layers.2.NC.1.norm2.bias, layers.2.NC.1.mlp.fc1.weight, layers.2.NC.1.mlp.fc1.bias, layers.2.NC.1.mlp.fc2.weight, layers.2.NC.1.mlp.fc2.bias, layers.2.NC.1.PCM.0.weight, layers.2.NC.1.PCM.0.bias, layers.2.NC.1.PCM.1.weight, layers.2.NC.1.PCM.1.bias, layers.2.NC.1.PCM.1.running_mean, layers.2.NC.1.PCM.1.running_var, layers.2.NC.1.PCM.1.num_batches_tracked, layers.2.NC.1.PCM.3.weight, layers.2.NC.1.PCM.3.bias, layers.2.NC.2.norm1.weight, layers.2.NC.2.norm1.bias, layers.2.NC.2.attn.qkv.bias, layers.2.NC.2.attn.qkv.weight, layers.2.NC.2.attn.proj.weight, layers.2.NC.2.attn.proj.bias, layers.2.NC.2.norm2.weight, layers.2.NC.2.norm2.bias, layers.2.NC.2.mlp.fc1.weight, layers.2.NC.2.mlp.fc1.bias, layers.2.NC.2.mlp.fc2.weight, layers.2.NC.2.mlp.fc2.bias, layers.2.NC.2.PCM.0.weight, layers.2.NC.2.PCM.0.bias, layers.2.NC.2.PCM.1.weight, layers.2.NC.2.PCM.1.bias, layers.2.NC.2.PCM.1.running_mean, layers.2.NC.2.PCM.1.running_var, layers.2.NC.2.PCM.1.num_batches_tracked, layers.2.NC.2.PCM.3.weight, layers.2.NC.2.PCM.3.bias, layers.2.NC.3.norm1.weight, layers.2.NC.3.norm1.bias, layers.2.NC.3.attn.qkv.bias, layers.2.NC.3.attn.qkv.weight, layers.2.NC.3.attn.proj.weight, layers.2.NC.3.attn.proj.bias, layers.2.NC.3.norm2.weight, layers.2.NC.3.norm2.bias, layers.2.NC.3.mlp.fc1.weight, layers.2.NC.3.mlp.fc1.bias, layers.2.NC.3.mlp.fc2.weight, layers.2.NC.3.mlp.fc2.bias, layers.2.NC.3.PCM.0.weight, layers.2.NC.3.PCM.0.bias, layers.2.NC.3.PCM.1.weight, layers.2.NC.3.PCM.1.bias, layers.2.NC.3.PCM.1.running_mean, layers.2.NC.3.PCM.1.running_var, layers.2.NC.3.PCM.1.num_batches_tracked, layers.2.NC.3.PCM.3.weight, layers.2.NC.3.PCM.3.bias, layers.2.NC.4.norm1.weight, layers.2.NC.4.norm1.bias, layers.2.NC.4.attn.qkv.bias, layers.2.NC.4.attn.qkv.weight, layers.2.NC.4.attn.proj.weight, layers.2.NC.4.attn.proj.bias, layers.2.NC.4.norm2.weight, layers.2.NC.4.norm2.bias, layers.2.NC.4.mlp.fc1.weight, layers.2.NC.4.mlp.fc1.bias, layers.2.NC.4.mlp.fc2.weight, layers.2.NC.4.mlp.fc2.bias, layers.2.NC.4.PCM.0.weight, layers.2.NC.4.PCM.0.bias, layers.2.NC.4.PCM.1.weight, layers.2.NC.4.PCM.1.bias, layers.2.NC.4.PCM.1.running_mean, layers.2.NC.4.PCM.1.running_var, layers.2.NC.4.PCM.1.num_batches_tracked, layers.2.NC.4.PCM.3.weight, layers.2.NC.4.PCM.3.bias, layers.2.NC.5.norm1.weight, layers.2.NC.5.norm1.bias, layers.2.NC.5.attn.qkv.bias, layers.2.NC.5.attn.qkv.weight, layers.2.NC.5.attn.proj.weight, layers.2.NC.5.attn.proj.bias, layers.2.NC.5.norm2.weight, layers.2.NC.5.norm2.bias, layers.2.NC.5.mlp.fc1.weight, layers.2.NC.5.mlp.fc1.bias, layers.2.NC.5.mlp.fc2.weight, layers.2.NC.5.mlp.fc2.bias, layers.2.NC.5.PCM.0.weight, layers.2.NC.5.PCM.0.bias, layers.2.NC.5.PCM.1.weight, layers.2.NC.5.PCM.1.bias, layers.2.NC.5.PCM.1.running_mean, layers.2.NC.5.PCM.1.running_var, layers.2.NC.5.PCM.1.num_batches_tracked, layers.2.NC.5.PCM.3.weight, layers.2.NC.5.PCM.3.bias, layers.2.NC.6.norm1.weight, layers.2.NC.6.norm1.bias, layers.2.NC.6.attn.qkv.bias, layers.2.NC.6.attn.qkv.weight, layers.2.NC.6.attn.proj.weight, layers.2.NC.6.attn.proj.bias, layers.2.NC.6.norm2.weight, layers.2.NC.6.norm2.bias, layers.2.NC.6.mlp.fc1.weight, layers.2.NC.6.mlp.fc1.bias, layers.2.NC.6.mlp.fc2.weight, layers.2.NC.6.mlp.fc2.bias, layers.2.NC.6.PCM.0.weight, layers.2.NC.6.PCM.0.bias, layers.2.NC.6.PCM.1.weight, layers.2.NC.6.PCM.1.bias, layers.2.NC.6.PCM.1.running_mean, layers.2.NC.6.PCM.1.running_var, layers.2.NC.6.PCM.1.num_batches_tracked, layers.2.NC.6.PCM.3.weight, layers.2.NC.6.PCM.3.bias, layers.2.NC.7.norm1.weight, layers.2.NC.7.norm1.bias, layers.2.NC.7.attn.qkv.bias, layers.2.NC.7.attn.qkv.weight, layers.2.NC.7.attn.proj.weight, layers.2.NC.7.attn.proj.bias, layers.2.NC.7.norm2.weight, layers.2.NC.7.norm2.bias, layers.2.NC.7.mlp.fc1.weight, layers.2.NC.7.mlp.fc1.bias, layers.2.NC.7.mlp.fc2.weight, layers.2.NC.7.mlp.fc2.bias, layers.2.NC.7.PCM.0.weight, layers.2.NC.7.PCM.0.bias, layers.2.NC.7.PCM.1.weight, layers.2.NC.7.PCM.1.bias, layers.2.NC.7.PCM.1.running_mean, layers.2.NC.7.PCM.1.running_var, layers.2.NC.7.PCM.1.num_batches_tracked, layers.2.NC.7.PCM.3.weight, layers.2.NC.7.PCM.3.bias, layers.2.NC.8.norm1.weight, layers.2.NC.8.norm1.bias, layers.2.NC.8.attn.qkv.bias, layers.2.NC.8.attn.qkv.weight, layers.2.NC.8.attn.proj.weight, layers.2.NC.8.attn.proj.bias, layers.2.NC.8.norm2.weight, layers.2.NC.8.norm2.bias, layers.2.NC.8.mlp.fc1.weight, layers.2.NC.8.mlp.fc1.bias, layers.2.NC.8.mlp.fc2.weight, layers.2.NC.8.mlp.fc2.bias, layers.2.NC.8.PCM.0.weight, layers.2.NC.8.PCM.0.bias, layers.2.NC.8.PCM.1.weight, layers.2.NC.8.PCM.1.bias, layers.2.NC.8.PCM.1.running_mean, layers.2.NC.8.PCM.1.running_var, layers.2.NC.8.PCM.1.num_batches_tracked, layers.2.NC.8.PCM.3.weight, layers.2.NC.8.PCM.3.bias, layers.2.NC.9.norm1.weight, layers.2.NC.9.norm1.bias, layers.2.NC.9.attn.qkv.bias, layers.2.NC.9.attn.qkv.weight, layers.2.NC.9.attn.proj.weight, layers.2.NC.9.attn.proj.bias, layers.2.NC.9.norm2.weight, layers.2.NC.9.norm2.bias, layers.2.NC.9.mlp.fc1.weight, layers.2.NC.9.mlp.fc1.bias, layers.2.NC.9.mlp.fc2.weight, layers.2.NC.9.mlp.fc2.bias, layers.2.NC.9.PCM.0.weight, layers.2.NC.9.PCM.0.bias, layers.2.NC.9.PCM.1.weight, layers.2.NC.9.PCM.1.bias, layers.2.NC.9.PCM.1.running_mean, layers.2.NC.9.PCM.1.running_var, layers.2.NC.9.PCM.1.num_batches_tracked, layers.2.NC.9.PCM.3.weight, layers.2.NC.9.PCM.3.bias, layers.2.NC.10.norm1.weight, layers.2.NC.10.norm1.bias, layers.2.NC.10.attn.qkv.bias, layers.2.NC.10.attn.qkv.weight, layers.2.NC.10.attn.proj.weight, layers.2.NC.10.attn.proj.bias, layers.2.NC.10.norm2.weight, layers.2.NC.10.norm2.bias, layers.2.NC.10.mlp.fc1.weight, layers.2.NC.10.mlp.fc1.bias, layers.2.NC.10.mlp.fc2.weight, layers.2.NC.10.mlp.fc2.bias, layers.2.NC.10.PCM.0.weight, layers.2.NC.10.PCM.0.bias, layers.2.NC.10.PCM.1.weight, layers.2.NC.10.PCM.1.bias, layers.2.NC.10.PCM.1.running_mean, layers.2.NC.10.PCM.1.running_var, layers.2.NC.10.PCM.1.num_batches_tracked, layers.2.NC.10.PCM.3.weight, layers.2.NC.10.PCM.3.bias, layers.2.NC.11.norm1.weight, layers.2.NC.11.norm1.bias, layers.2.NC.11.attn.qkv.bias, layers.2.NC.11.attn.qkv.weight, layers.2.NC.11.attn.proj.weight, layers.2.NC.11.attn.proj.bias, layers.2.NC.11.norm2.weight, layers.2.NC.11.norm2.bias, layers.2.NC.11.mlp.fc1.weight, layers.2.NC.11.mlp.fc1.bias, layers.2.NC.11.mlp.fc2.weight, layers.2.NC.11.mlp.fc2.bias, layers.2.NC.11.PCM.0.weight, layers.2.NC.11.PCM.0.bias, layers.2.NC.11.PCM.1.weight, layers.2.NC.11.PCM.1.bias, layers.2.NC.11.PCM.1.running_mean, layers.2.NC.11.PCM.1.running_var, layers.2.NC.11.PCM.1.num_batches_tracked, layers.2.NC.11.PCM.3.weight, layers.2.NC.11.PCM.3.bias, head.weight, head.bias, fc_norm.weight, fc_norm.bias

missing keys in source state_dict: pos_embed, patch_embed.proj.weight, patch_embed.proj.bias, blocks.0.norm1.weight, blocks.0.norm1.bias, blocks.0.attn.rel_pos_h, blocks.0.attn.rel_pos_w, blocks.0.attn.qkv.weight, blocks.0.attn.qkv.bias, blocks.0.attn.proj.weight, blocks.0.attn.proj.bias, blocks.0.norm2.weight, blocks.0.norm2.bias, blocks.0.mlp.fc1.weight, blocks.0.mlp.fc1.bias, blocks.0.mlp.fc2.weight, blocks.0.mlp.fc2.bias, blocks.0.PCM.0.weight, blocks.0.PCM.0.bias, blocks.0.PCM.1.weight, blocks.0.PCM.1.bias, blocks.0.PCM.1.running_mean, blocks.0.PCM.1.running_var, blocks.0.PCM.3.weight, blocks.0.PCM.3.bias, blocks.1.norm1.weight, blocks.1.norm1.bias, blocks.1.attn.rel_pos_h, blocks.1.attn.rel_pos_w, blocks.1.attn.qkv.weight, blocks.1.attn.qkv.bias, blocks.1.attn.proj.weight, blocks.1.attn.proj.bias, blocks.1.norm2.weight, blocks.1.norm2.bias, blocks.1.mlp.fc1.weight, blocks.1.mlp.fc1.bias, blocks.1.mlp.fc2.weight, blocks.1.mlp.fc2.bias, blocks.1.PCM.0.weight, blocks.1.PCM.0.bias, blocks.1.PCM.1.weight, blocks.1.PCM.1.bias, blocks.1.PCM.1.running_mean, blocks.1.PCM.1.running_var, blocks.1.PCM.3.weight, blocks.1.PCM.3.bias, blocks.2.norm1.weight, blocks.2.norm1.bias, blocks.2.attn.q_bias, blocks.2.attn.v_bias, blocks.2.attn.rel_pos_h, blocks.2.attn.rel_pos_w, blocks.2.attn.qkv.weight, blocks.2.attn.proj.weight, blocks.2.attn.proj.bias, blocks.2.norm2.weight, blocks.2.norm2.bias, blocks.2.mlp.fc1.weight, blocks.2.mlp.fc1.bias, blocks.2.mlp.fc2.weight, blocks.2.mlp.fc2.bias, blocks.2.PCM.0.weight, blocks.2.PCM.0.bias, blocks.2.PCM.1.weight, blocks.2.PCM.1.bias, blocks.2.PCM.1.running_mean, blocks.2.PCM.1.running_var, blocks.2.PCM.3.weight, blocks.2.PCM.3.bias, blocks.3.norm1.weight, blocks.3.norm1.bias, blocks.3.attn.rel_pos_h, blocks.3.attn.rel_pos_w, blocks.3.attn.qkv.weight, blocks.3.attn.qkv.bias, blocks.3.attn.proj.weight, blocks.3.attn.proj.bias, blocks.3.norm2.weight, blocks.3.norm2.bias, blocks.3.mlp.fc1.weight, blocks.3.mlp.fc1.bias, blocks.3.mlp.fc2.weight, blocks.3.mlp.fc2.bias, blocks.3.PCM.0.weight, blocks.3.PCM.0.bias, blocks.3.PCM.1.weight, blocks.3.PCM.1.bias, blocks.3.PCM.1.running_mean, blocks.3.PCM.1.running_var, blocks.3.PCM.3.weight, blocks.3.PCM.3.bias, blocks.4.norm1.weight, blocks.4.norm1.bias, blocks.4.attn.rel_pos_h, blocks.4.attn.rel_pos_w, blocks.4.attn.qkv.weight, blocks.4.attn.qkv.bias, blocks.4.attn.proj.weight, blocks.4.attn.proj.bias, blocks.4.norm2.weight, blocks.4.norm2.bias, blocks.4.mlp.fc1.weight, blocks.4.mlp.fc1.bias, blocks.4.mlp.fc2.weight, blocks.4.mlp.fc2.bias, blocks.4.PCM.0.weight, blocks.4.PCM.0.bias, blocks.4.PCM.1.weight, blocks.4.PCM.1.bias, blocks.4.PCM.1.running_mean, blocks.4.PCM.1.running_var, blocks.4.PCM.3.weight, blocks.4.PCM.3.bias, blocks.5.norm1.weight, blocks.5.norm1.bias, blocks.5.attn.q_bias, blocks.5.attn.v_bias, blocks.5.attn.rel_pos_h, blocks.5.attn.rel_pos_w, blocks.5.attn.qkv.weight, blocks.5.attn.proj.weight, blocks.5.attn.proj.bias, blocks.5.norm2.weight, blocks.5.norm2.bias, blocks.5.mlp.fc1.weight, blocks.5.mlp.fc1.bias, blocks.5.mlp.fc2.weight, blocks.5.mlp.fc2.bias, blocks.5.PCM.0.weight, blocks.5.PCM.0.bias, blocks.5.PCM.1.weight, blocks.5.PCM.1.bias, blocks.5.PCM.1.running_mean, blocks.5.PCM.1.running_var, blocks.5.PCM.3.weight, blocks.5.PCM.3.bias, blocks.6.norm1.weight, blocks.6.norm1.bias, blocks.6.attn.rel_pos_h, blocks.6.attn.rel_pos_w, blocks.6.attn.qkv.weight, blocks.6.attn.qkv.bias, blocks.6.attn.proj.weight, blocks.6.attn.proj.bias, blocks.6.norm2.weight, blocks.6.norm2.bias, blocks.6.mlp.fc1.weight, blocks.6.mlp.fc1.bias, blocks.6.mlp.fc2.weight, blocks.6.mlp.fc2.bias, blocks.6.PCM.0.weight, blocks.6.PCM.0.bias, blocks.6.PCM.1.weight, blocks.6.PCM.1.bias, blocks.6.PCM.1.running_mean, blocks.6.PCM.1.running_var, blocks.6.PCM.3.weight, blocks.6.PCM.3.bias, blocks.7.norm1.weight, blocks.7.norm1.bias, blocks.7.attn.rel_pos_h, blocks.7.attn.rel_pos_w, blocks.7.attn.qkv.weight, blocks.7.attn.qkv.bias, blocks.7.attn.proj.weight, blocks.7.attn.proj.bias, blocks.7.norm2.weight, blocks.7.norm2.bias, blocks.7.mlp.fc1.weight, blocks.7.mlp.fc1.bias, blocks.7.mlp.fc2.weight, blocks.7.mlp.fc2.bias, blocks.7.PCM.0.weight, blocks.7.PCM.0.bias, blocks.7.PCM.1.weight, blocks.7.PCM.1.bias, blocks.7.PCM.1.running_mean, blocks.7.PCM.1.running_var, blocks.7.PCM.3.weight, blocks.7.PCM.3.bias, blocks.8.norm1.weight, blocks.8.norm1.bias, blocks.8.attn.q_bias, blocks.8.attn.v_bias, blocks.8.attn.rel_pos_h, blocks.8.attn.rel_pos_w, blocks.8.attn.qkv.weight, blocks.8.attn.proj.weight, blocks.8.attn.proj.bias, blocks.8.norm2.weight, blocks.8.norm2.bias, blocks.8.mlp.fc1.weight, blocks.8.mlp.fc1.bias, blocks.8.mlp.fc2.weight, blocks.8.mlp.fc2.bias, blocks.8.PCM.0.weight, blocks.8.PCM.0.bias, blocks.8.PCM.1.weight, blocks.8.PCM.1.bias, blocks.8.PCM.1.running_mean, blocks.8.PCM.1.running_var, blocks.8.PCM.3.weight, blocks.8.PCM.3.bias, blocks.9.norm1.weight, blocks.9.norm1.bias, blocks.9.attn.rel_pos_h, blocks.9.attn.rel_pos_w, blocks.9.attn.qkv.weight, blocks.9.attn.qkv.bias, blocks.9.attn.proj.weight, blocks.9.attn.proj.bias, blocks.9.norm2.weight, blocks.9.norm2.bias, blocks.9.mlp.fc1.weight, blocks.9.mlp.fc1.bias, blocks.9.mlp.fc2.weight, blocks.9.mlp.fc2.bias, blocks.9.PCM.0.weight, blocks.9.PCM.0.bias, blocks.9.PCM.1.weight, blocks.9.PCM.1.bias, blocks.9.PCM.1.running_mean, blocks.9.PCM.1.running_var, blocks.9.PCM.3.weight, blocks.9.PCM.3.bias, blocks.10.norm1.weight, blocks.10.norm1.bias, blocks.10.attn.rel_pos_h, blocks.10.attn.rel_pos_w, blocks.10.attn.qkv.weight, blocks.10.attn.qkv.bias, blocks.10.attn.proj.weight, blocks.10.attn.proj.bias, blocks.10.norm2.weight, blocks.10.norm2.bias, blocks.10.mlp.fc1.weight, blocks.10.mlp.fc1.bias, blocks.10.mlp.fc2.weight, blocks.10.mlp.fc2.bias, blocks.10.PCM.0.weight, blocks.10.PCM.0.bias, blocks.10.PCM.1.weight, blocks.10.PCM.1.bias, blocks.10.PCM.1.running_mean, blocks.10.PCM.1.running_var, blocks.10.PCM.3.weight, blocks.10.PCM.3.bias, blocks.11.norm1.weight, blocks.11.norm1.bias, blocks.11.attn.q_bias, blocks.11.attn.v_bias, blocks.11.attn.rel_pos_h, blocks.11.attn.rel_pos_w, blocks.11.attn.qkv.weight, blocks.11.attn.proj.weight, blocks.11.attn.proj.bias, blocks.11.norm2.weight, blocks.11.norm2.bias, blocks.11.mlp.fc1.weight, blocks.11.mlp.fc1.bias, blocks.11.mlp.fc2.weight, blocks.11.mlp.fc2.bias, blocks.11.PCM.0.weight, blocks.11.PCM.0.bias, blocks.11.PCM.1.weight, blocks.11.PCM.1.bias, blocks.11.PCM.1.running_mean, blocks.11.PCM.1.running_var, blocks.11.PCM.3.weight, blocks.11.PCM.3.bias, norm.weight, norm.bias, fpn1.0.weight, fpn1.0.bias, fpn1.1.ln.weight, fpn1.1.ln.bias, fpn1.3.weight, fpn1.3.bias, fpn2.0.weight, fpn2.0.bias
`