lucidrains soft-moe-pytorch issues - Githubissues

lucidrains / soft-moe-pytorch

Implementation of Soft MoE, proposed by Brain's Vision team, in Pytorch

MIT License

239 stars 8 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Here are some questions about soft MoE

#9 t5862755 opened 4 months ago
0
Early onset of Inf error on a certain self supervised learning method in distributed training mode

#8 swarajnanda2021 closed 6 months ago
1
Unable to Set num_slots Directly

#7 afiff2 closed 6 months ago
1
Still updating and maintainancing?

#6 Shengguang-Zhou opened 9 months ago
0
have a question

#5 daixiangzi closed 10 months ago
5
Not receiving grads with cpu_besides?

#4 conceptofmind closed 4 months ago
26
error:tImplementedError: Could not run 'aten::_amp_foreach_non_finite_check_and_unscale_' with arguments from the 'CPU' backend.

#3 daixiangzi opened 11 months ago
24
Does the number of slots * the number of experts have to be the same as the length of the token?

#2 ZQpengyu opened 11 months ago
1