Tag: Branch: Tree: 73768ef6e3

master

Yolo_v2

Yolo_v2_tested

Yolo_v3

darknet_yolo_v3

darknet_yolo_v3_optimal

darknet_yolo_v4_pre

yolov4

Branches Tags

${ item.name }

Create tag ${ searchTerm }

Create branch ${ searchTerm }

from '73768ef6e3'

${ noResults }

Commit Graph

1834 Commits (73768ef6e3ee3647e198022300a9e0a7e751ef5b)

 

All Branches   

Search

Author	SHA1	Message	Date
AlexeyAB	73768ef6e3	Added [net] max_chart_loss=50 for scale y-axis of Loss-chart	5 years ago
AlexeyAB	bef28445e5	Fixed mish gradient	5 years ago
AlexeyAB	2500278784	compile fix	5 years ago
AlexeyAB	0acd6990d8	Minor speedup of Mish/Swish (if train==0 then don't write input of activation to the memory)	5 years ago
AlexeyAB	9cd44eea95	Minor fix	5 years ago
AlexeyAB	0c7305cf63	Minor fix	5 years ago
AlexeyAB	74134e5389	Fixed gradient_centralization for Grouped-Conv	5 years ago
AlexeyAB	9f8b1e31a8	Minor fix	5 years ago
AlexeyAB	88a39b31d3	Added param [yolo] resize=2.0 for resizing image [1/2 - 2x] without changing aspect ratio (while jitter=0.3 changes aspect ratio)	5 years ago
AlexeyAB	f5105342be	Added Gradient Centralization: [convolutional] grad_centr=1	5 years ago
AlexeyAB	6ad485c0c1	Fixed cutmix=1 error message.	5 years ago
Alexey	da3de2bdf9	Update Readme.md	5 years ago
AlexeyAB	8fede92885	Fixed CUDA Out of memory error during training	5 years ago
AlexeyAB	b5870c12d6	./darknet detector valid uses letter_box=1 from cfg-file [net] letter_box=1	5 years ago
AlexeyAB	da050367a8	Another Demo speed fix. Also added flag -dontdraw_bbox for OpenDataCam	5 years ago
AlexeyAB	424632f336	Fixed typo weights_normalizion -> weights_normalization	5 years ago
AlexeyAB	ccafff912b	Demo speed fixed	5 years ago
AlexeyAB	1f60ca57ad	Minor fix	5 years ago
Alexey	1d95870468	Update Readme.md	5 years ago
AlexeyAB	0fffb686fd	Added param [yolo] objectness_smooth=1	5 years ago
AlexeyAB	08bc0c9373	Fixed output of time_remaining for multi-GPU	5 years ago
Alexey	d4c10b0262	Merge pull request #5400 from ali-babaei/master ... [ fix ] - Not printing Iteration info in log file	5 years ago
Alexey	97df6ac2b4	Merge pull request #5479 from cenit/dev/cenit/appveyor_removal ... remove appveyor CI	5 years ago
Stefano Sinigardi	502fd51cd1	remove appveyor badge	5 years ago
Stefano Sinigardi	cba6153a41	remove appveyor CI	5 years ago
Alexey	37b62bc08c	Merge pull request #5417 from cenit/dev/cenit/ci_updates ... ci upgrades and half precision for cc >=7.0	5 years ago
Stefano Sinigardi	861192df19	shorten github actions names	5 years ago
AlexeyAB	6cbb75d10b	Minor fix for yolov4-custom.cfg	5 years ago
Stefano Sinigardi	551e57dc8b	fix linux make and add windows cygwin on github actions	5 years ago
Stefano Sinigardi	ac7436dccd	ported build configs from circleCI	5 years ago
Stefano Sinigardi	24ad413a79	add mingw and cuda on github actions, restored cuda 10.2 on appveyor as a test	5 years ago
Stefano Sinigardi	f2786ba19d	expand github action ci coverage, leave only cygwin on appveyor. Next step will be cuda on github actions	5 years ago
Stefano Sinigardi	72b3712df8	force release mode	5 years ago
Stefano Sinigardi	a993060d09	remove unnecessary system dependencies on github actions	5 years ago
Stefano Sinigardi	c3dc2c4a0d	use Release as default build type	5 years ago
Stefano Sinigardi	234218b961	add GitHub Actions	5 years ago
Stefano Sinigardi	a3f6c2b58e	merge with master	5 years ago
AlexeyAB	f14054ec2b	Fixed Cmake for CUDA CC 7.0, 7.2, 7.5	5 years ago
Stefano Sinigardi	98e4d246e1	make cuda version parametric on appveyor	5 years ago
Stefano Sinigardi	6afe964d4f	upgrade CUDA, cmake and VS on CI	5 years ago
Stefano Sinigardi	1c64c2d892	enable cc70,cc72,cc75 for half_precision, add message for omp missing on macOS	5 years ago
Stefano Sinigardi	fe879f3baf	simplify cmake due to minimum project requirement already gone to 3.12	5 years ago
Stefano Sinigardi	be49ab30cd	omp support on macOS	5 years ago
Ali Babaei	885abb5046	Merge pull request #1 from ali-babaei/ali-babaei-patch-1 ... [ fix ] - Not printing Iteration info in log file	5 years ago
Ali Babaei	1c1275fe88	[ fix ] - Not printing Iteration info in log file ... When we redirect output of training into log files, while the net.cudnn_half is True, iteration information line, which contains "avg loss" is missed. Using \n and fflush() fixed the issue. Each one of them may be enough to fix the issue.	5 years ago
AlexeyAB	36c73c5b9e	Some explanations for training	5 years ago
AlexeyAB	6d38218a04	Minor fix	5 years ago
AlexeyAB	65506eb04a	Minor fix	5 years ago
AlexeyAB	83b3d46ede	Fixed minor bug with closing ./darknet detector demo ...	5 years ago
Alexey	8e272fa868	Update Readme.md	5 years ago