Unable to load the state dict model

vision
1

I have saved the model using this code:
torch.save(best_model.state_dict(),'./leaf_Classsification_GPU_CutMix_EfficientNet-B4.pt')

then when I tried to load the model using this:
model = Net()

model.load_state_dict(torch.load("leaf_Classsification_GPU_CutMix_EfficientNet-B4.pt"))
model.eval()

it is giving me error:
AttributeError: ‘_IncompatibleKeys’ object has no attribute ‘eval’

Please suggest.

THIS IS THE MODEL WHICH IS TO BE LOADED for evaluation

2

Can you show me the complete code of your Net class?

3

Yes why not , here’s the link to my notebook .

and here’s my model class:

class EfficientNet_b4(nn.Module):
    def __init__(self):
        super(EfficientNet_b3, self).__init__()
        self.model = efficientnet_pytorch.EfficientNet.from_pretrained('efficientnet-b4')
        self.dropout = nn.Dropout(0.1)
        self.final_layer = nn.Linear(1792 , 5)
        
    def forward(self, inputs):
        batch_size, _, _, _ = inputs.shape
        
        x = self.model.extract_features(inputs)

        # Pooling and final linear layer
        x = self.model._avg_pooling(x)
        
        x = F.adaptive_avg_pool2d(x, 1).reshape(batch_size, -1)
        outputs = self.final_layer(self.dropout(x))

        return outputs
    
model = efficientnet_pytorch.EfficientNet.from_pretrained('efficientnet-b4')