feifeibear Follow

feifeibear Follow

💭

I may be slow to respond.

feifeibear feifeibear

💭

I may be slow to respond.

Senior Software Engineer at Wechat @Tencent.

Follow

68 followers · 31 following · 92

WeChat, Tencent
Beijing, China
Sign in to view email
https://fangjiarui.github.io/

Highlights

Arctic Code Vault Contributor

Organizations

Pinned

SWCaffe

A Deep Learning Framework customized for Sunway TaihuLight

C++ 30 18
swDNN

a highly-efficient library for deep neural networks based on Sunway TaihuLight supercomputer.

Roff 8 6
swGEMM

A highly efficient library for GEMM operations on Sunway TaihuLight

C 8 1
TurboTransformers

Forked from Tencent/TurboTransformers

a fast and user-friendly tool for transformer inference on CPU and GPU

C++ 1

812 contributions in the last year

Activity overview

Contributed to Tencent/TurboTransformers, feifeibear/test-ci, TurboNLP/Translate-Demo and 5 other repositories

Contribution activity

August 2020

feifeibear has no activity yet for this period.

July 2020

Created a pull request in Tencent/TurboTransformers that received 2 comments

Develop

BERT cpp output return value.

+55 −51 • 2 comments

Jiaruifang/fix onnxrt docker Jul 28
Jiaruifang/fix onnxrt return bug Jul 27
Jiaruifang/albert example Jul 24
Jiaruifang/onnxrt gpt Jul 24
Jiaruifang/onnxrt as turbo Jul 23
Jiaruifang/cpu variable len benchmark Jul 22
Rewrite cpu dev dockerfile. Solve mkl slow problems. Jul 22
Jiaruifang/random benchmark Jul 21
Jiaruifang/onnx benchmark opt Jul 20
Jiaruifang/fix allocator bug Jul 19
uses cub Jul 18
Develop Jul 15
More information in bert_example.py Jul 14
readme dockerhub image to lastest Jul 14
Jiaruifang/gpu docker v0.3.2 Jul 13
Develop Jul 13
Jiaruifang/v0.3.1 polish docker Jul 13
Develop Jul 10
Jiaruifang/multi model benchmark Jul 10
Jiaruifang/reberta tunr Jul 9
Jiaruifang/roberta Jul 9
Jiaruifang/albert model Jul 7
Jiaruifang/polish develop Jul 1
Jiaruifang/polish albert Jul 1
Jiaruifang/albert Jul 1

Fix/bert example bug Jul 30
fix bert model cpp output Jul 30
Add Qbertmodel Jul 24
Develop: Add QBertLayer Jul 23
clean up and fix typos Jul 22
Develop: Add QBertIntermediate and QBertOutput (#103) Jul 22
Fix gpu_benchmark & embedding kernel profile Jul 21
Add a preliminary test script that demonstrates a quantized bert layer Jul 17
move embedding to kernels Jul 17
add unittest for AddBias and fix some typos Jul 14
fix typo Jul 10

Created an issue in Tencent/TurboTransformers that received 4 comments

ONNXRT can not be applied in Albert

/opt/conda/lib/python3.7/site-packages/torch/onnx/utils.py:738: UserWarning: ONNX export failed on ATen operator einsum because torch.onnx.symbolic…

4 comments

MKL is slow in Turbo Jul 22
Refactor MultiheadAttention and other Layers Jul 22

You can’t perform that action at this time.