运行demo用cenn.pth预训练模型显示Expected 512, got 64

[whistbibi]

我改了在cpu上运行和去掉covert（‘L’）
model = model.cuda()

print('loading pretrained model from %s' % model_path)

if params.multi_gpu:

model = torch.nn.DataParallel(model)
checkpoint =torch.load(model_path, map_location=torch.device('cpu'))
model.load_state_dict(checkpoint,False)

converter = utils.strLabelConverter(params.alphabet)

transformer = dataset.resizeNormalize((100, 32))

image = Image.open(image_path).convert('L')

image = Image.open(image_path)
image = transformer(image)
if torch.cuda.is_available():
image = image.cuda()

image = image.view(1, *image.size())
image = Variable(image)

log：

`CRNN(
(cnn): Sequential(
(conv0): Conv2d(3, 64, kernel_size=(5, 5), stride=(2, 2), padding=(2, 2))
(relu0): ReLU(inplace=True)
(pooling0): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)
(conv1): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(relu1): ReLU(inplace=True)
(conv2): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(batchnorm2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(relu2): ReLU(inplace=True)
(conv3): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(pooling2): MaxPool2d(kernel_size=(2, 2), stride=(2, 1), padding=(0, 1), dilation=1, ceil_mode=False)
(conv4): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(batchnorm4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(conv5): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(pooling3): MaxPool2d(kernel_size=(2, 2), stride=(2, 1), padding=(0, 1), dilation=1, ceil_mode=False)
(conv6): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(batchnorm6): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
(rnn): Sequential(
(0): BidirectionalLSTM(
(rnn): LSTM(512, 256, bidirectional=True)
(embedding): Linear(in_features=512, out_features=256, bias=True)
)
(1): BidirectionalLSTM(
(rnn): LSTM(256, 256, bidirectional=True)
(embedding): Linear(in_features=512, out_features=204, bias=True)
)
)
)
loading pretrained model from ./path/to/crnn.pth

Traceback (most recent call last):
File "demo.py", line 45, in
preds = model(image)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 140, in forward
return self.module(*inputs, **kwargs)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/dm10801/B1_tinyAI_project/202110-cruiser/1.1creat_pytorch_module/models/crnn.py", line 87, in forward
output = self.rnn(conv)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/modules/container.py", line 100, in forward
input = module(input)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/dm10801/B1_tinyAI_project/202110-cruiser/1.1creat_pytorch_module/models/crnn.py", line 14, in forward
recurrent, _ = self.rnn(input)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/modules/rnn.py", line 556, in forward
self.check_forward_args(input, hx, batch_sizes)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/modules/rnn.py", line 508, in check_forward_args
self.check_input(input, batch_sizes)
File "/usrhome/jingli/.conda/envs/202110-crnn-License-plate-env/lib/python3.6/site-packages/torch/nn/modules/rnn.py", line 159, in check_input
self.input_size, input.size(-1)))
RuntimeError: input.size(-1) must be equal to input_size. Expected 512, got 64`