greenhouse/data/hyp.scratch.yaml
Glenn Jocher 76807fae71
YOLOv5 Forward Compatibility Update (#1569)
* YOLOv5 forward compatibility update

* add data dir

* ci test yolov3

* update build_targets()

* update build_targets()

* update build_targets()

* update yolov3-spp.yaml

* add yolov3-tiny.yaml

* add yolov3-tiny.yaml

* Update yolov3-tiny.yaml

* thop bug fix

* Detection() device bug fix

* Use torchvision.ops.nms()

* Remove redundant download mirror

* CI tests with yolov3-tiny

* Update README.md

* Synch train and test iou_thresh

* update requirements.txt

* Cat apriori autolabels

* Confusion matrix

* Autosplit

* Autosplit

* Update README.md

* AP no plot

* Update caching

* Update caching

* Caching bug fix

* --image-weights bug fix

* datasets bug fix

* mosaic plots bug fix

* plot_study

* boxes.max()

* boxes.max()

* boxes.max()

* boxes.max()

* boxes.max()

* boxes.max()

* update

* Update README

* Update README

* Update README.md

* Update README.md

* results png

* Update README

* Targets scaling bug fix

* update plot_study

* update plot_study

* update plot_study

* update plot_study

* Targets scaling bug fix

* Finish Readme.md

* Finish Readme.md

* Finish Readme.md

* Update README.md

* Creado con Colaboratory
2020-11-26 20:24:00 +01:00

34 lines
1.5 KiB
YAML

# Hyperparameters for COCO training from scratch
# python train.py --batch 40 --cfg yolov5m.yaml --weights '' --data coco.yaml --img 640 --epochs 300
# See tutorials for hyperparameter evolution https://github.com/ultralytics/yolov5#tutorials
lr0: 0.01 # initial learning rate (SGD=1E-2, Adam=1E-3)
lrf: 0.2 # final OneCycleLR learning rate (lr0 * lrf)
momentum: 0.937 # SGD momentum/Adam beta1
weight_decay: 0.0005 # optimizer weight decay 5e-4
warmup_epochs: 3.0 # warmup epochs (fractions ok)
warmup_momentum: 0.8 # warmup initial momentum
warmup_bias_lr: 0.1 # warmup initial bias lr
box: 0.05 # box loss gain
cls: 0.5 # cls loss gain
cls_pw: 1.0 # cls BCELoss positive_weight
obj: 1.0 # obj loss gain (scale with pixels)
obj_pw: 1.0 # obj BCELoss positive_weight
iou_t: 0.20 # IoU training threshold
anchor_t: 4.0 # anchor-multiple threshold
# anchors: 3 # anchors per output layer (0 to ignore)
fl_gamma: 0.0 # focal loss gamma (efficientDet default gamma=1.5)
hsv_h: 0.015 # image HSV-Hue augmentation (fraction)
hsv_s: 0.7 # image HSV-Saturation augmentation (fraction)
hsv_v: 0.4 # image HSV-Value augmentation (fraction)
degrees: 0.0 # image rotation (+/- deg)
translate: 0.1 # image translation (+/- fraction)
scale: 0.5 # image scale (+/- gain)
shear: 0.0 # image shear (+/- deg)
perspective: 0.0 # image perspective (+/- fraction), range 0-0.001
flipud: 0.0 # image flip up-down (probability)
fliplr: 0.5 # image flip left-right (probability)
mosaic: 1.0 # image mosaic (probability)
mixup: 0.0 # image mixup (probability)