models/VITClassifier.py

import os
import torch
import torch.nn as nn
import torch.nn.init as init
from .vit import ViT

class ViTLargeClassifier(nn.Module):
    def __init__(self, num_classes: int = 14, image_size=224, patch_size=16):
        super(ViTLargeClassifier, self).__init__()
        
        # 初始化ViT模型
        self.vit = ViT(
            image_size=image_size,
            patch_size=patch_size,
            num_classes=num_classes,
            dim=1024,
            depth=24,
            heads=16,
            mlp_dim=4096,
            dropout=0.1,
            emb_dropout=0.1
        )
        
        # 初始化权重
        if not self.load():
            for m in self.modules():
                if isinstance(m, nn.Linear):
                    init.xavier_normal_(m.weight)
                    if m.bias is not None:
                        init.zeros_(m.bias)
                elif isinstance(m, nn.LayerNorm):
                    init.ones_(m.weight)
                    init.zeros_(m.bias)
    
    def forward(self, x):
        return self.vit(x)
    
    def load(self, filename: str = None) -> bool:
        if filename is None:
            current_work_dir = os.path.dirname(__file__)
            filename = os.path.join(current_work_dir, "best_pth", "ViTLargeClassifier.pth")
        if not os.path.exists(filename):
            print("Model file does not exist.")
            return False
        self.load_state_dict(torch.load(filename))
        print("Model loaded successfully.")
        return True