~speedprog/mtg/mtg_card_detector.git - Our Gitblit

speedprog / mtg/mtg_card_detector

first | « prev | next »

Fixed multi-GPU training for Tensor Cores

AlexeyAB

2018-03-09 880cf187d87c904f5fe574802ecff99118643f2d

2018-03-09

Fixed multi-GPU training for Tensor Cores

blob@ 880cf1

commitdiff | diff to current

2018-02-25

Added support for Tensor Cores CC >= 7.0 (V100). For FP16/32 (mixed precisi...

blob@ cad4d1

commitdiff | diff to current

2018-02-23

Updated to CUDA 9.1. And fixed no_gpu dependecies.

blob@ cd2bde

commitdiff | diff to current

2018-02-22

blob@ 6332ea

commitdiff | diff to current

2018-02-22

Added __float2half_rn() and __half2float()

blob@ b2b575

commitdiff | diff to current

2018-02-22

Use half_float16 instead of float32 if defined both CUDNN and CUDNN_HALF. U...

blob@ dda993

commitdiff | diff to current

2017-07-13

blob@ 992041

commitdiff | diff to current

2017-01-16

Fixed behavior if missing library cudnn.lib

blob@ d7a30a

commitdiff | diff to current

2017-01-15

Fixed behavior if missing library cudnn.lib

blob@ 3b9afd

commitdiff | diff to current

2016-11-25

:vegan: :charizard:

blob@ 75fe60

commitdiff | diff to current

2016-11-05

new font strategy

blob@ c7a700

commitdiff | diff to current

2016-10-26

blob@ 352ae7

commitdiff | diff to current

2016-09-20

better multigpu

blob@ 73f7aa

commitdiff | diff to current

2016-09-12

good chance I didn't break anything

blob@ 5c067d

commitdiff | diff to current

2016-09-01

updates and things

blob@ 8f1b4e

commitdiff | diff to current

2016-06-23

blob@ afb8b4

commitdiff | diff to current

2016-06-19

no mean on input binarization

blob@ 08c7cf

commitdiff | diff to current

2016-06-14

blob@ 8322a5

commitdiff | diff to current

2016-06-10

blob@ 729ce4

commitdiff | diff to current

2016-06-02

hope i didn't break anything

blob@ ec3d05

commitdiff | diff to current

2016-05-13

blob@ 13209d

commitdiff | diff to current

2016-05-06

so much need to commit

blob@ c7b10c

commitdiff | diff to current

2016-03-16

blob@ cff59b

commitdiff | diff to current

2016-03-14

Go

blob@ d1965b

commitdiff | diff to current

2016-02-29

blob@ 16d06e

commitdiff | diff to current

2016-01-28

blob@ 913d35

commitdiff | diff to current

2015-12-08

fixed darknet, stuff

blob@ 892923

commitdiff | diff to current

2015-12-08

Faster batch normalization

blob@ c27388

commitdiff | diff to current

2015-11-16

Fix for cuda 7.5

blob@ 0f7f28

commitdiff | diff to current

2015-11-04

CVPR Experiments

blob@ 8fd18a

commitdiff | diff to current

2015-07-21

Changes to make routing work better

blob@ d00f0a

commitdiff | diff to current

2015-03-30

blob@ 6553b3

commitdiff | diff to current

2015-03-27

Fixed im2col mistake >< face#palm

blob@ d7d7da

commitdiff | diff to current

2015-03-25

smaller gridsize in bias

blob@ e92f7d

commitdiff | diff to current

2015-03-23

going to break stuff

blob@ 7100de

commitdiff | diff to current

2015-03-22

Subdivisions for batches

blob@ 664c5d

commitdiff | diff to current

2015-03-21

using caffe's im2col, it's so much better\!

blob@ 9d4181

commitdiff | diff to current

2015-03-21

gonna change im2col

blob@ 4af116

commitdiff | diff to current

2015-03-12

refactoring and added DARK ZONE

blob@ dcb000

commitdiff | diff to current

2015-02-11

Detection is back, baby\!

blob@ 0f6458

commitdiff | diff to current

2015-02-09

Generalizing conv layer so deconv is easier

blob@ 979d02

commitdiff | diff to current

2015-02-04

Stable place to commit

blob@ bfffad

commitdiff | diff to current

2015-01-27

Bias updates bug fix

blob@ 153705

commitdiff | diff to current

2015-01-23

blob@ 809f92

commitdiff | diff to current

first | « prev | next »