Bienqq
/
car-detection-bayes
2
1
Fork 0
Code Issues Pull Requests Releases Wiki Activity

updates

This commit is contained in:
Glenn Jocher 2020-03-10 12:17:23 -07:00
parent d55dbc1f29
commit 7a83574022
1 changed files with 6 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
train.py
View File

@ -36,10 +36,10 @@ hyp = {'giou': 3.54, # giou loss gain
'hsv_h': 0.0138, # image HSV-Hue augmentation (fraction) 'hsv_h': 0.0138, # image HSV-Hue augmentation (fraction)
'hsv_s': 0.678, # image HSV-Saturation augmentation (fraction) 'hsv_s': 0.678, # image HSV-Saturation augmentation (fraction)
'hsv_v': 0.36, # image HSV-Value augmentation (fraction) 'hsv_v': 0.36, # image HSV-Value augmentation (fraction)
'degrees': 1.98, # image rotation (+/- deg) 'degrees': 1.98 * 0, # image rotation (+/- deg)
'translate': 0.05, # image translation (+/- fraction) 'translate': 0.05 * 0, # image translation (+/- fraction)
'scale': 0.05, # image scale (+/- gain) 'scale': 0.05 * 0, # image scale (+/- gain)
'shear': 0.641} # image shear (+/- deg) 'shear': 0.641 * 0} # image shear (+/- deg)
# Overwrite hyp with hyp*.txt (optional) # Overwrite hyp with hyp*.txt (optional)
f = glob.glob('hyp*.txt') f = glob.glob('hyp*.txt')
@ -197,7 +197,7 @@ def train():
# Start training # Start training
nb = len(dataloader) nb = len(dataloader)
prebias = start_epoch == 0 prebias = False # start_epoch == 0
model.nc = nc # attach number of classes to model model.nc = nc # attach number of classes to model
model.arc = opt.arc # attach yolo architecture model.arc = opt.arc # attach yolo architecture
model.hyp = hyp # attach hyperparameters to model model.hyp = hyp # attach hyperparameters to model
@ -211,7 +211,7 @@ def train():
print('Starting training for %g epochs...' % epochs) print('Starting training for %g epochs...' % epochs)
for epoch in range(start_epoch, epochs): # epoch ------------------------------------------------------------------ for epoch in range(start_epoch, epochs): # epoch ------------------------------------------------------------------
model.train() model.train()
model.gr = 1 - (1 + math.cos(min(epoch * 2, epochs) * math.pi / epochs)) / 2 # GIoU <-> 1.0 loss ratio model.gr = 1 - (1 + math.cos(min(epoch * 1, epochs) * math.pi / epochs)) / 2 # GIoU <-> 1.0 loss ratio
# Prebias # Prebias
if prebias: if prebias: