profile
viewpoint
If you are wondering where the data of this site comes from, please visit https://api.github.com/users/laekov/events. GitMemory does not store any data, but only uses NGINX to cache data for a period of time. The idea behind GitMemory is simply to give users a better reading experience.

laekov/fastmoe 253

A fast MoE impl for PyTorch

haoxizhong/TUOJ 10

Let's discover a new world. — Edit

laekov/acejudge 9

A simple terminal OI/ACM answer checker on Linux/UNIX

laekov/autohpl 4

SGD tuning HPL

laekov/dgebc 4

Distributed Genetic Evolutional Box2d Cars

Bakser/CST62_FOP 3

FOP2016 course codes of CST62,Tsinghua University

laekov/auto-ribao 2

Automatically fill the Ribao chart of THU during the COVID-2019

laekov/capiano 2

Virtual piano using finger capture through cameras

laekov/emnlp2017-relation-extraction 2

Context-Aware Representations for Knowledge Base Relation Extraction

startedrwschubert/kalman-laika

started time in 7 hours

startedMKorostoff/1-pixel-wealth

started time in 7 hours

startedtermoshtt/link_cuda_kernel

started time in 9 hours

startedvuejs/vue

started time in 12 hours

startedjardicc/alchemist

started time in 12 hours

startedlaekov/fastmoe

started time in 13 hours

startedlaekov/fastmoe

started time in 13 hours

startedcommaai/laika

started time in 14 hours

startedcommaai/openpilot

started time in 14 hours

startedtibold/svg-explorer-extension

started time in 15 hours

startedmpetroff/kindle-weather-display

started time in 17 hours

starteddottedmag/x2x

started time in 2 days

startedgaogaotiantian/viztracer

started time in 2 days

startedmpetroff/kindle-weather-display

started time in 2 days

startedthu-scc/homepage

started time in 2 days

startedSU-CISEC/gpu-ntt

started time in 2 days

startedant-design/ant-design

started time in 2 days

startedant-design/ant-design

started time in 2 days

startedant-design/ant-design

started time in 2 days

issue closedlaekov/fastmoe

magic number (256) in CUDA functions

There is a magic number (256) in Both in CUDA functions moe_cuda_local_scatter_impl and moe_cuda_local_gather_impl. I cannot understand what it means and not sure if it's a potential bug in fastmoe. Is it related to the parameters of hardwares?

Related codes: batch_scatter_kernel<scalar_t> <<<batch_size, 256, 0, smgr->stream(0)>>>(in_feat, d_pos, input, input_buf);

closed time in 2 days

zjujh1995

issue commentlaekov/fastmoe

magic number (256) in CUDA functions

You can find the meaning of this number in any cuda programming tutorial, e..g. https://developer.nvidia.com/blog/easy-introduction-cuda-c-and-c/. The number defines how many device threads execute the kernel in parallel.

zjujh1995

comment created time in 2 days

startedcostela/wesher

started time in 2 days

startedprivacybot-berkeley/privacybot

started time in 2 days

startedNVIDIA/thrust

started time in 2 days

startedNVIDIA/jitify

started time in 2 days

startedNVIDIA/cutlass

started time in 2 days

startedestk/log4rs

started time in 3 days

startedhellodword/wechat-feeds

started time in 3 days