-
There are a few independent implementations for sequential recommenders (sequence-aware, session-based, and so on) such as [slientGe's repository](https://github.com/slientGe/Sequential_Recommendation…
-
Good evening.
#### Describe the bug
I found a weird thing in the MTTKRP operation.
The order of factor matrices in MTTKRP is not the same with a formulation of CP-ALS.
However, the fit (factoriz…
-
Dear Ivana,
thank you for great works [1,3,4] among many others. I was wondering whether one could find the best performing hyper parameters for Hyper along with the number of parameters of it.
…
-
This issue is for tracking work in progress for UVM removal
===========================================
# UVM Free Tools
- [x] CellTools - mauro done
- [x] Discretization - kyungjoo, mauro, nate…
-
#### Is your feature request related to a problem? Please describe.
Currently it looks like we simply take the standard SVD and then the absolute value of every element, for non-negative CP and Tuc…
-
## 🐛 Bug
## To Reproduce
Steps to reproduce the behavior:
1. Add identity pruning to a module's weight to reparameterize it
1. Call `register_forward_pre_hook` to register additional hoo…
-
## 🐛 Bug
Running the tests I get 2 failing tests in test_torch where NaNs are produced
## To Reproduce
Steps to reproduce the behavior:
1. Build with `PYTORCH_BUILD_VERSION=1.7.0 PYTORCH_B…
-
Hi, thanks for sharing the code, I have some questions about Equation 2 in this paper.
1. In the original ComplEx paper of ICML 2016, the scoring function is Re(). While in Equation 2 of your paper…
-
Very interesting work! A quick question, can we understand this method as a generalization of factorization machine (https://cseweb.ucsd.edu/classes/fa17/cse291-b/reading/Rendle2010FM.pdf)? it appears…
-
@apaszke convinced me that Dex can do named tensors even in its current form and provide type checking. It's pretty close. Here's a prototype that implements the current attention formulation using na…
srush updated
3 years ago