【扒代码】ope模块的模型结构

# 查看模型结构（每个层的结构） 
print(ope)

# 查看模型名称和大小
for name, param in ope.named_parameters():
    print(name, param.size())
# 查看模块的参数数量
param_count = sum(p.numel() for p in ope.parameters())
print(f"Total number of parameters: {param_count}")

'''

output_size: torch.Size([3, 27, 4, 256])

iterative_adaptation.layers.0.norm1.weight torch.Size([256])

iterative_adaptation.layers.0.norm1.bias torch.Size([256])

iterative_adaptation.layers.0.norm2.weight torch.Size([256])

iterative_adaptation.layers.0.norm2.bias torch.Size([256])

iterative_adaptation.layers.0.norm3.weight torch.Size([256])

iterative_adaptation.layers.0.norm3.bias torch.Size([256])

iterative_adaptation.layers.0.self_attn.in_proj_weight torch.Size([768, 256])

iterative_adaptation.layers.0.self_attn.in_proj_bias torch.Size([768])

iterative_adaptation.layers.0.self_attn.out_proj.weight torch.Size([256, 256])

iterative_adaptation.layers.0.self_attn.out_proj.bias torch.Size([256])

iterative_adaptation.layers.0.enc_dec_attn.in_proj_weight torch.Size([768, 256])

iterative_adaptation.layers.0.enc_dec_attn.in_proj_bias torch.Size([768])

iterative_adaptation.layers.0.enc_dec_attn.out_proj.weight torch.Size([256, 256])

iterative_adaptation.layers.0.enc_dec_attn.out_proj.bias torch.Size([256])

iterative_adaptation.layers.0.mlp.linear1.weight torch.Size([2048, 256])

iterative_adaptation.layers.0.mlp.linear1.bias torch.Size([2048])

iterative_adaptation.layers.0.mlp.linear2.weight torch.Size([256, 2048])

iterative_adaptation.layers.0.mlp.linear2.bias torch.Size([256])

---

iterative_adaptation.layers.1.norm1.weight torch.Size([256])

iterative_adaptation.layers.1.norm1.bias torch.Size([256])

iterative_adaptation.layers.1.norm2.weight torch.Size([256])

iterative_adaptation.layers.1.norm2.bias torch.Size([256])

iterative_adaptation.layers.1.norm3.weight torch.Size([256])

iterative_adaptation.layers.1.norm3.bias torch.Size([256])

iterative_adaptation.layers.1.self_attn.in_proj_weight torch.Size([768, 256])

iterative_adaptation.layers.1.self_attn.in_proj_bias torch.Size([768])

iterative_adaptation.layers.1.self_attn.out_proj.weight torch.Size([256, 256])

iterative_adaptation.layers.1.self_attn.out_proj.bias torch.Size([256])

iterative_adaptation.layers.1.enc_dec_attn.in_proj_weight torch.Size([768, 256])

iterative_adaptation.layers.1.enc_dec_attn.in_proj_bias torch.Size([768])

iterative_adaptation.layers.1.enc_dec_attn.out_proj.weight torch.Size([256, 256])

iterative_adaptation.layers.1.enc_dec_attn.out_proj.bias torch.Size([256])

iterative_adaptation.layers.1.mlp.linear1.weight torch.Size([2048, 256])

iterative_adaptation.layers.1.mlp.linear1.bias torch.Size([2048])

iterative_adaptation.layers.1.mlp.linear2.weight torch.Size([256, 2048])

iterative_adaptation.layers.1.mlp.linear2.bias torch.Size([256])

---

iterative_adaptation.layers.2.norm1.weight torch.Size([256])

iterative_adaptation.layers.2.norm1.bias torch.Size([256])

iterative_adaptation.layers.2.norm2.weight torch.Size([256])

iterative_adaptation.layers.2.norm2.bias torch.Size([256])

iterative_adaptation.layers.2.norm3.weight torch.Size([256])

iterative_adaptation.layers.2.norm3.bias torch.Size([256])

iterative_adaptation.layers.2.self_attn.in_proj_weight torch.Size([768, 256])

iterative_adaptation.layers.2.self_attn.in_proj_bias torch.Size([768])

iterative_adaptation.layers.2.self_attn.out_proj.weight torch.Size([256, 256])

iterative_adaptation.layers.2.self_attn.out_proj.bias torch.Size([256])

iterative_adaptation.layers.2.enc_dec_attn.in_proj_weight torch.Size([768, 256])

iterative_adaptation.layers.2.enc_dec_attn.in_proj_bias torch.Size([768])

iterative_adaptation.layers.2.enc_dec_attn.out_proj.weight torch.Size([256, 256])

iterative_adaptation.layers.2.enc_dec_attn.out_proj.bias torch.Size([256])

iterative_adaptation.layers.2.mlp.linear1.weight torch.Size([2048, 256])

iterative_adaptation.layers.2.mlp.linear1.bias torch.Size([2048])

iterative_adaptation.layers.2.mlp.linear2.weight torch.Size([256, 2048])

iterative_adaptation.layers.2.mlp.linear2.bias torch.Size([256])

iterative_adaptation.norm.weight torch.Size([256])

iterative_adaptation.norm.bias torch.Size([256])

---

shape_or_objectness.0.weight torch.Size([64, 2])

shape_or_objectness.0.bias torch.Size([64])

shape_or_objectness.2.weight torch.Size([256, 64])

shape_or_objectness.2.bias torch.Size([256])

shape_or_objectness.4.weight torch.Size([2304, 256])

shape_or_objectness.4.bias torch.Size([2304])

---

Total number of parameters: 5345728

'''