search

cemoody / topicsne

t-SNE experiments in pytorch
473 stars 85 forks source link

'indices' to have scalar type Long error #5

Open eashaankumar opened 5 years ago

eashaankumar commented 5 years ago

This code does not work when I try to run the run.py script. It throws this error:

1083 2 2 Traceback (most recent call last): File "run.py", line 48, in <module> wrap.fit(pij, i, j) File "C:\github\topicsne\wrapper.py", line 44, in fit loss = self.model(*datas) File "C:\github\topicsne\vtsne.py", line 67, in __call__ return self.forward(*args) File "C:\github\topicsne\vtsne.py", line 57, in forward xi, _ = self.sample_logits(i) File "C:\github\topicsne\vtsne.py", line 46, in sample_logits return self.reparametrize(self.logits_mu(i), self.logits_lv(i)) File "C:\Users\easha\Anaconda3\envs\pytorch\lib\site-packages\torch\nn\modules\module.py", line 489, in __call__ result = self.forward(*input, **kwargs) File "C:\Users\easha\Anaconda3\envs\pytorch\lib\site-packages\torch\nn\modules\sparse.py", line 118, in forward self.norm_type, self.scale_grad_by_freq, self.sparse) File "C:\Users\easha\Anaconda3\envs\pytorch\lib\site-packages\torch\nn\functional.py", line 1454, in embedding return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse) RuntimeError: Expected tensor for argument #1 'indices' to have scalar type Long; but got torch.cuda.IntTensor instead (while checking arguments for embedding)

I tried to mess around with the suggested data types, but that doesn't seem to solve the issue.

miaoxiaodaiblack commented 2 years ago

Have you solved the problem?

eashaankumar commented 2 years ago

Hi,

This was a while ago. I do not remember if I did.