cumf / cumf_sgd Goto Github PK
View Code? Open in Web Editor NEWCUDA Matrix Factorization Library with Stochastic Gradient Descent (SGD)
CUDA Matrix Factorization Library with Stochastic Gradient Descent (SGD)
Hello
I want to run cumfsgd on Yahoo! Music dataset.
How should I set the parameters to get the result?
Can you please add a license to your repository ?
Thanks
Are there any alternative links?
How and where do I extract X and Theta after convergence? Are these being written to a file on the host device?
Hi, I think there is a bug in the 89-99 lines of cumf_sgd/singleGPU/sgd_k128_kernel_hogwild_warp32.h:
//update
//only works for k=blockDim.x=128
p[base_p + lane_id + 0] = __float2half(tmp_p1 + tmp_lrate*(ruv*tmp_q1 - lambda_p*tmp_p1));
q[base_q + lane_id + 0] = __float2half(tmp_q1 + tmp_lrate*(ruv*tmp_p1 - lambda_q*tmp_q1));
p[base_p + lane_id + 32] = __float2half(tmp_p2 + tmp_lrate*(ruv*tmp_q2 - lambda_p*tmp_p2));
q[base_q + lane_id + 32] = __float2half(tmp_q2 + tmp_lrate*(ruv*tmp_p2 - lambda_q*tmp_q2));
p[base_p + lane_id + 64] = __float2half(tmp_p3 + tmp_lrate*(ruv*tmp_q3 - lambda_p*tmp_p3));
q[base_q + lane_id + 64] = __float2half(tmp_q3 + tmp_lrate*(ruv*tmp_p3 - lambda_q*tmp_q3));
p[base_p + lane_id + 96] = __float2half(tmp_p4 + tmp_lrate*(ruv*tmp_q4 - lambda_p*tmp_p4));
q[base_q + lane_id + 96] = __float2half(tmp_q4 + tmp_lrate*(ruv*tmp_p4 - lambda_q*tmp_q4));
Multiple workers(blocks) or warps may update the same base_p or base_q so that there will be write conflicts. This procedure should use atomic operations.
Hi,
I'm doing some unrelated research, resurrecting a project from a couple years ago. It relies on the netflix_mm and netflix_mme files, which seem to have completely disappeared from the Internet. Your README does mention these files as being in your google drive link, but I can't seem to find them (you have netflix_train and netflix_test, but these appear to be in a different format).
Do you, by any chance, have the netflix_mm and netflix_mme files, and would it be at all possible for you to (temporarily) upload them somewhere where I could download them? It would be much appreciated.
Thanks in advance!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.