Pytorch изображения декодер дает несоответствие размера

Question

Ниже приведен мой код

import os
import argparse

import numpy as np
from scipy.misc import imread, imresize, imsave

import torch
from torch.autograd import Variable
# device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
parser = argparse.ArgumentParser()
parser.add_argument('--model', required=True, type=str, help='path to model')
parser.add_argument('--input', required=True, type=str, help='input codes')
parser.add_argument('--output', default='.', type=str, help='output folder')
parser.add_argument('--cuda', action='store_true', help='enables cuda')
parser.add_argument(
'--iterations', type=int, default=16, help='unroll iterations')
args = parser.parse_args()

content = np.load(args.input)
codes = np.unpackbits(content['codes'])
codes = np.reshape(codes, content['shape']).astype(np.float32) * 2 - 1

codes = torch.from_numpy(codes)
iters, batch_size, channels, height, width = codes.size()
height = height * 16
# print(height)
width = width * 16
# print(width)

with torch.no_grad():
     codes = Variable(codes, requires_grad=True)

import network

decoder = network.DecoderCell()
decoder.eval()

decoder.load_state_dict(torch.load(args.model), strict=False)
# decoder.load_state_dict(torch.load(args.model, map_location={'cuda:0': 'cpu'}))
# decoder.load_state_dict(torch.load(args.model, map_location=lambda storage, location: storage))

decoder_h_1 = (Variable(
    torch.zeros(batch_size, 512, height // 16, width // 16), requires_grad=True),
            Variable(
               torch.zeros(batch_size, 512, height // 16, width // 16),
               requires_grad=True))
decoder_h_2 = (Variable(
   torch.zeros(batch_size, 512, height // 8, width // 8), requires_grad=True),
           Variable(
               torch.zeros(batch_size, 512, height // 8, width // 8),
               requires_grad=True))
decoder_h_3 = (Variable(
   torch.zeros(batch_size, 256, height // 4, width // 4), requires_grad=True),
           Variable(
               torch.zeros(batch_size, 256, height // 4, width // 4),
               requires_grad=True))
decoder_h_4 = (Variable(
   torch.zeros(batch_size, 128, height // 2, width // 2), requires_grad=True),
           Variable(
               torch.zeros(batch_size, 128, height // 2, width // 2),
               requires_grad=True))

if args.cuda:
   decoder = decoder.cuda()

   codes = codes.cuda()

    decoder_h_1 = (decoder_h_1[0].cuda(), decoder_h_1[1].cuda())
    decoder_h_2 = (decoder_h_2[0].cuda(), decoder_h_2[1].cuda())
    decoder_h_3 = (decoder_h_3[0].cuda(), decoder_h_3[1].cuda())
    decoder_h_4 = (decoder_h_4[0].cuda(), decoder_h_4[1].cuda())

image = torch.zeros(1, 3, height, width) + 0.5
for iters in range(min(args.iterations, codes.size(0))):

    output, decoder_h_1, decoder_h_2, decoder_h_3, decoder_h_4 = decoder(
    codes[iters], decoder_h_1, decoder_h_2, decoder_h_3, decoder_h_4)
    image = image + output.data.cpu()

    imsave(os.path.join(args.output, '{:02d}.png'.format(iters)),np.squeeze(image.numpy().clip(0, 1) * 255.0).astype(np.uint8).transpose(1, 2, 0))

После я получаю сообщение об ошибке типа несоответствие размера

Traceback (most recent call last):
File "decoder.py", line 38, in <module>
decoder.load_state_dict(torch.load(args.model), strict=False)
File "C:\Users\User1\Anaconda4\lib\site-packages\torch\nn\modules\module.py", line 719, in load_state_dict
self.__class__.__name__, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for DecoderCell:
    size mismatch for rnn1.conv_ih.weight: copying a param of torch.Size([2048, 512, 3, 3]) from 
checkpoint, where the shape is torch.Size([1024, 64, 3, 3]) in current model.
    size mismatch for rnn1.conv_hh.weight: copying a param of torch.Size([2048, 512, 1, 1]) from 
checkpoint, where the shape is torch.Size([1024, 256, 1, 1]) in current model.
    size mismatch for rnn2.conv_ih.weight: copying a param of torch.Size([2048, 128, 3, 3]) from 
checkpoint, where the shape is torch.Size([2048, 256, 3, 3]) in current model.
    size mismatch for rnn3.conv_ih.weight: copying a param of torch.Size([1024, 128, 3, 3]) from 
checkpoint, where the shape is torch.Size([2048, 512, 3, 3]) in current model.
    size mismatch for rnn3.conv_hh.weight: copying a param of torch.Size([1024, 256, 3, 3]) from 
checkpoint, where the shape is torch.Size([2048, 512, 1, 1]) in current model.

Я получаю вышеуказанную ошибку, Как устранить ошибку, потому что я тренируюсь изображения для PNG и до сих пор получаю вышеуказанную ошибку. есть ли конкретное c решение для этой ошибки.

Я использую PyTorch 0.4. python 3,6 Использование размера изображения с 64 X 64