a97f350461
Non-output layer freeze in train.py ( #1333 )
...
Freeze layers that aren't of type YOLOLayer and that aren't the conv layers preceeding them
2020-06-22 13:15:40 -07:00
10dc08f91b
revert value of gs back to 32(from 64) ( #1317 )
2020-06-19 09:54:57 -07:00
c78d49f190
check_file() update from yolov5
2020-06-15 12:25:48 -07:00
002884ae5e
multi_label burnin addition
2020-05-21 14:40:45 -07:00
da40084b37
burnin update
2020-05-17 21:03:36 -07:00
316d99c377
yolov5 regress updates to yolov3
2020-05-17 15:19:33 -07:00
c94019f159
iglob bug fix
2020-05-17 14:31:14 -07:00
37bd5490ef
iglob file-search improvements
2020-05-16 22:25:21 -07:00
0cf88f046d
hyp evolution bug fix #1160
2020-05-12 09:53:13 -07:00
d405959893
cleanup
2020-05-04 13:33:34 -07:00
fb1b5e09b2
faster and more informative training plots ( #1114 )
...
* faster and more informative training plots
* Update utils.py
Looks good. Needs pep8 linting, I'll do that in PyCharm later once PR is in.
* Update test.py
* Update train.py
f for the tb descriptor lets us plot several batches, i.e. to allow us to change L292 to 'if ni < 3' for 3 examples.
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2020-04-30 13:37:04 -07:00
f1d73a29e5
Optimizer group report
2020-04-30 12:26:02 -07:00
d62d68929c
cleanup
2020-04-29 12:00:30 -07:00
9f88f5cc21
cleanup
2020-04-29 11:34:59 -07:00
37cbe89ef0
test/train jpg for png
2020-04-28 13:45:27 -07:00
8521c3cff9
cleanup
2020-04-27 15:22:36 -07:00
5a8efa5c1d
auto --accumulate
2020-04-23 14:32:28 -07:00
b3dfd89878
scheduler resume bug fix
2020-04-23 10:35:08 -07:00
345d65d18f
updated train default img_size 320-640
2020-04-22 14:32:05 -07:00
7cbac5a3ea
train.py iou_t to 0.20
2020-04-22 11:34:34 -07:00
77b3829d56
check_git_status() to train.py
2020-04-22 11:02:09 -07:00
a5bd0fe567
tensorboard comment=opt.name
2020-04-21 12:19:29 -07:00
be3f322375
Tensorboard out of try, iou_t to 0.10
2020-04-20 09:57:15 -07:00
c3edf8daf4
move image size report
2020-04-15 22:03:51 -07:00
510eadcfa5
Apex and 'git pull' suggestions
2020-04-15 12:54:56 -07:00
f5a2682a81
image sizes report
2020-04-14 12:02:08 -07:00
763cdd5ae2
detailed image sizes report
2020-04-14 11:51:19 -07:00
1038b0d269
multi-scale update
2020-04-12 18:22:54 -07:00
ada2958105
Fix argparse string escapes in train.py. ( #1045 )
2020-04-12 10:00:50 -07:00
6e19245dc8
auto strip optimizer from best.pt after training
2020-04-09 19:53:29 -07:00
d1601ae0f3
training updates
2020-04-08 21:34:34 -07:00
4120ac3aa6
training updates
2020-04-08 21:01:58 -07:00
97780cfdb4
parameterize grid size
2020-04-08 10:14:33 -07:00
4fa3fd2df3
tensorboard notice
2020-04-06 16:33:23 -07:00
05ae6e8499
tensorboard/focal loss reporting update
2020-04-06 15:45:18 -07:00
c7f93bae40
update coco-tuned hyp['cls'] to current dataset
2020-04-06 10:58:07 -07:00
e81a152a92
tensorboard notice and model verbose option
2020-04-05 13:35:58 -07:00
41246aa042
tensorboard updates
2020-04-04 19:34:39 -07:00
9155ef3f4f
burnin merged with prebias
2020-04-02 14:08:21 -07:00
992e0d7cb4
default test --conf to 0.001
2020-03-31 14:36:25 -07:00
b2d9f1898f
burnin lr ramp 300 iterations
2020-03-30 19:27:42 -07:00
eb151a881e
NMS and test batch_size updates
2020-03-29 20:41:32 -07:00
c6b59a0e8a
LR schedule to 0.05 min
2020-03-29 13:29:06 -07:00
9c5e76b93d
EMA implemented by default
2020-03-29 13:14:54 -07:00
20b2671de0
cleanup ( #963 )
...
* cleanup
cleanup
* Update train.py
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2020-03-26 14:46:17 -07:00
448c4a6e1f
Remove deprecated --arc architecture options, implement --arc default for all cases
2020-03-16 20:46:25 -07:00
851c9b9883
EMA class updates
2020-03-14 17:34:13 -07:00
d91469a516
EMA class updates
2020-03-14 17:33:29 -07:00
9ce4ec48a7
model.info() method implemented
2020-03-14 16:46:54 -07:00
b89cc396af
EMA class updates
2020-03-14 16:23:14 -07:00
Oulbacha Reda
FuLin
Glenn Jocher
Josh Veitch-Michaelis
Timothy M. Shead
Yonghye Kwon