dist = tag_mean.unsqueeze(1) - tag_mean.unsqueeze(2)

[cqray1990]

dist = tag_mean.unsqueeze(1) - tag_mean.unsqueeze(2)

when the code run here,it raised : dist = tag_mean.unsqueeze(1) - tag_mean.unsqueeze(2) RuntimeError: Dimension out of range (expected to be in range of [-2, 1], but got 2)

can you help me? Thank you!

[berther]

replace the two lines tag0 = tag0[:].squeeze() tag1 = tag1[:].squeeze() with tag0 = torch.squeeze(tag0,dim=2) tag1 = torch.squeeze(tag1,dim=2)

dist = tag_mean.unsqueeze(1) - tag_mean.unsqueeze(2)

when the code run here,it raised : dist = tag_mean.unsqueeze(1) - tag_mean.unsqueeze(2) RuntimeError: Dimension out of range (expected to be in range of [-2, 1], but got 2)

can you help me? Thank you!

replace the two lines tag0 = tag0[:].squeeze() tag1 = tag1[:].squeeze() with tag0 = torch.squeeze(tag0,dim=2) tag1 = torch.squeeze(tag1,dim=2)

[cqray1990]

tag0 is two dimension，does it works?

[cqray1990]

tag0 is two dimension，does it works?@berther

[berther]

tag0 is two dimension，does it works?@berther

After the two lines,If tag0 and tag1 are two dimension,it works.Is your batchsize equal to 1?

[cqray1990]

Yes，i set batch_size=1

[cqray1990]

Yes，i set batch_size=1,@berther

[cqray1990]

but i get another problem,even i set batch_size=1,it also raise RuntimeError: CUDA out of memory. Tried to allocate 6.75 MiB (GPU 0; 3.94 GiB total capacity; 1.50 GiB already allocated; 23.00 MiB free; 49.10 MiB cached),it is wired??