RuntimeError: mat1 and mat2 shapes cannot be multiplied (1x16384 and 4096x8192)

[T-xu]

请问下作者和大家，训练的时候有碰到这个问题吗，训练代码一直没有调通，配置文件只修改了图片的路径。不知道哪里出的问题。期待解答，谢谢。

Traceback (most recent call last): File "gfpgan/train.py", line 11, in train_pipeline(root_path) File "/data/mingyang/tx/BasicSR-master/basicsr/train.py", line 169, in train_pipeline model.optimize_parameters(current_iter) File "/home/mingyang/anaconda3/envs/tx/lib/python3.7/site-packages/gfpgan/models/gfpgan_model.py", line 305, in optimize_parameters self.output, out_rgbs = self.net_g(self.lq, return_rgb=True) File "/home/mingyang/anaconda3/envs/tx/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl return forward_call(*input, **kwargs) File "/home/mingyang/anaconda3/envs/tx/lib/python3.7/site-packages/torch/nn/parallel/distributed.py", line 799, in forward output = self.module(*inputs[0], **kwargs[0]) File "/home/mingyang/anaconda3/envs/tx/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl return forward_call(*input, **kwargs) File "/home/mingyang/anaconda3/envs/tx/lib/python3.7/site-packages/gfpgan/archs/gfpganv1_arch.py", line 355, in forward style_code = self.final_linear(feat.view(feat.size(0), -1)) File "/home/mingyang/anaconda3/envs/tx/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl return forward_call(*input, **kwargs) File "/data/mingyang/tx/BasicSR-master/basicsr/archs/stylegan2_arch.py", line 174, in forward out = F.linear(x, self.weight * self.scale, bias=bias) File "/home/mingyang/anaconda3/envs/tx/lib/python3.7/site-packages/torch/nn/functional.py", line 1847, in linear return torch._C._nn.linear(input, weight, bias) RuntimeError: mat1 and mat2 shapes cannot be multiplied (1x16384 and 4096x8192) ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: 1) local_rank: 0 (pid: 48829) of binary: /home/mingyang/anaconda3/envs/tx/bin/python

[alexliyang]

这个是直接用10241024的图像训练了吧，resieze到512512可以开始训练，但是是否应该resize呢？

[wangqilongcv]

这个是直接用1024_1024的图像训练了吧，resieze到512_512可以开始训练，但是是否应该resize呢？

请问最后是怎么处理的，resize后是否对最后效果有影响

[sunjian2015]

这个是直接用1024_1024的图像训练了吧，resieze到512_512可以开始训练，但是是否应该resize呢？

请问最后是怎么处理的，resize后是否对最后效果有影响

请问你是怎么处理的呢？我resize后是可以跑通的。