nickgreenquist cu2rec issues

nickgreenquist / cu2rec

CUDA Implementation of Parallel Matrix Factorization Algorithm for Recommender Systems

GNU Affero General Public License v3.0

13 stars 1 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

One does not simply prepare for final submission

#52 nickgreenquist closed 5 years ago
0
Change test to validation, add 3-way dataset split

#51 nickgreenquist opened 5 years ago
0
Add all variables in config.h to actual config files that are read in

#50 nickgreenquist opened 5 years ago
0
Check the number of actual updates being done

#49 dorukkilitcioglu opened 5 years ago
0
P and Q factors to registers, implement early bird wins the gradient

#48 nickgreenquist closed 5 years ago
0
Look at ways of parallelizing over factors

#47 dorukkilitcioglu opened 5 years ago
0
Use Unsigned Integers instead of floats at all

#46 nickgreenquist opened 5 years ago
0
Drop floats to half precision

#45 nickgreenquist opened 5 years ago
0
Add Atomic write to Q_target and item_bias_target

#44 nickgreenquist closed 5 years ago
1
delete recommendations folder

#43 nickgreenquist closed 5 years ago
0
put train and test losses in seperate streams

#42 nickgreenquist opened 5 years ago
0
Netflix

#41 dorukkilitcioglu closed 5 years ago
0
MF cpu version

#40 nickgreenquist closed 5 years ago
0
add n_threads to config, clean up minor other things

#39 nickgreenquist closed 5 years ago
0
Add every constant value to Config

#38 nickgreenquist closed 5 years ago
0
Expirement with larger block dimensions

#37 nickgreenquist closed 5 years ago
1
make cu2rec item agnostic

#36 nickgreenquist closed 5 years ago
0
Handle matrices with missing users

#35 dorukkilitcioglu closed 5 years ago
1
Add unit test for total loss kernel

#34 dorukkilitcioglu closed 5 years ago
1
Total loss

#33 dorukkilitcioglu closed 5 years ago
0
Working code for 20m ratings test/train split

#32 nickgreenquist closed 5 years ago
0
compute single loss in sgd, move prediction code to devicde function,…

#31 nickgreenquist closed 5 years ago
0
Add Train/Test split

#30 nickgreenquist closed 5 years ago
0
add test/train split

#29 nickgreenquist closed 5 years ago
0
Implement Training Loop Termination using Patience

#28 nickgreenquist closed 5 years ago
0
Implement adaptive Learning Rate

#27 nickgreenquist closed 5 years ago
0
Comments & documentation

#26 dorukkilitcioglu closed 5 years ago
2
Partial fit

#25 dorukkilitcioglu closed 5 years ago
0
Remove local vars in kernels that are used only once

#24 nickgreenquist closed 5 years ago
0
Put user bias into shared memory in SGD

#23 dorukkilitcioglu closed 5 years ago
1
Training should depend on the config

#22 dorukkilitcioglu closed 5 years ago
1
Turn config into JSON

#21 dorukkilitcioglu opened 5 years ago
0
Remove utils.cuh

#20 dorukkilitcioglu closed 5 years ago
0
Merge config branch

#19 dorukkilitcioglu closed 5 years ago
0
Partial fit

#18 dorukkilitcioglu closed 5 years ago
2
P and Q regularization

#17 dorukkilitcioglu closed 6 years ago
1
refactor read_csv to util, add code to write trained components to files

#16 nickgreenquist closed 6 years ago
0
Create config struct for hyperparameters

#15 dorukkilitcioglu closed 5 years ago
5
Add Global Bias

#14 nickgreenquist closed 6 years ago
0
add item and user bias to loss, sgd, and trainining code

#13 nickgreenquist closed 6 years ago
0
Multi-GPU Grid Search for Hyperparameters

#12 nickgreenquist opened 6 years ago
0
Performance Comparisons

#11 dorukkilitcioglu closed 5 years ago
1
K-Fold validation?

#10 dorukkilitcioglu closed 5 years ago
1
Minibatch gradient descent

#9 dorukkilitcioglu opened 6 years ago
3
Normalize gradient updates by the number of training examples

#8 dorukkilitcioglu closed 5 years ago
2
Get recommendations for a new user

#7 dorukkilitcioglu closed 5 years ago
2
Add Dot Product usable in Kernels

#6 nickgreenquist closed 5 years ago
1
Output Trained P and Q to files

#5 nickgreenquist closed 5 years ago
1
Add User/Item Bias to Loss and SGD

#4 nickgreenquist closed 6 years ago
0
Add training loop and test

#3 dorukkilitcioglu closed 5 years ago
2