AntoineH/smart_augmentation
1
0
Fork 0
mirror of https://github.com/AntoineHX/smart_augmentation.git synced 2025-05-04 04:00:46 +02:00
Code Issues Projects Releases Packages Wiki Activity

MaJ example simple + Suppression Nesterov momentum

Browse source
This commit is contained in:
Harle, Antoine (Contracteur) 2020-02-17 17:32:54 -05:00
parent b170af076f
commit d53b385c43
4 changed files with 22 additions and 129 deletions
Download patch file Download diff file Expand all files Collapse all files

45
higher/smart_aug/benchmark.py
View file

@ -19,7 +19,7 @@ optim_param={
'lr':1e-1, #1e-2/1e-1 (ResNet) 'lr':1e-1, #1e-2/1e-1 (ResNet)
'momentum':0.9, #0.9 'momentum':0.9, #0.9
'decay':0.0005, #0.0005 'decay':0.0005, #0.0005
'nesterov':True, 'nesterov':False, #False (True: Bad behavior w/ Data_aug)
'scheduler':'cosine', #None, 'cosine', 'multiStep', 'exponential' 'scheduler':'cosine', #None, 'cosine', 'multiStep', 'exponential'
} }
} }
@ -30,44 +30,6 @@ epochs= 200
dataug_epoch_start=0 dataug_epoch_start=0
nb_run= 3 nb_run= 3
# Use available TF (see transformations.py)
'''
tf_names = [
## Geometric TF ##
'Identity',
'FlipUD',
'FlipLR',
'Rotate',
'TranslateX',
'TranslateY',
'ShearX',
'ShearY',
## Color TF (Expect image in the range of [0, 1]) ##
'Contrast',
'Color',
'Brightness',
'Sharpness',
'Posterize',
'Solarize', #=>Image entre [0,1] #Pas opti pour des batch
## Bad Tranformations ##
# Bad Geometric TF #
#'BShearX',
#'BShearY',
#'BTranslateX-',
#'BTranslateX-',
#'BTranslateY',
#'BTranslateY-',
#'BadContrast',
#'BadBrightness',
#'Random',
#'RandBlend'
]
tf_dict = {k: TF.TF_dict[k] for k in tf_names}
'''
tf_config='../config/base_tf_config.json' tf_config='../config/base_tf_config.json'
TF_loader=TF_loader() TF_loader=TF_loader()
tf_dict, tf_ignore_mag =TF_loader.load_TF_dict(tf_config) tf_dict, tf_ignore_mag =TF_loader.load_TF_dict(tf_config)
@ -91,7 +53,7 @@ if __name__ == "__main__":
### Benchmark ### ### Benchmark ###
#''' #'''
n_inner_iter = 1 n_inner_iter = 1
dist_mix = [0.5]#[0.5, 1.0] dist_mix = [0.5, 1.0]
N_seq_TF= [3, 4] N_seq_TF= [3, 4]
mag_setup = [(False, False)] #[(True, True), (False, False)] #(FxSh, Independant) mag_setup = [(False, False)] #[(True, True), (False, False)] #(FxSh, Independant)
@ -117,7 +79,8 @@ if __name__ == "__main__":
mix_dist=dist, mix_dist=dist,
fixed_prob=False, fixed_prob=False,
fixed_mag=m_setup[0], fixed_mag=m_setup[0],
shared_mag=m_setup[1]), shared_mag=m_setup[1],
TF_ignore_mag=tf_ignore_mag),
model).to(device) model).to(device)
else: else:
aug_model = Augmented_model(RandAug(TF_dict=tf_dict, N_TF=n_tf), model).to(device) aug_model = Augmented_model(RandAug(TF_dict=tf_dict, N_TF=n_tf), model).to(device)

38
higher/smart_aug/smart_aug_example.py
View file

@ -2,31 +2,12 @@
""" """
from model import * from LeNet import *
from dataug import * from dataug import *
from train_utils import * from train_utils import *
# Use available TF (see transformations.py) tf_config='../config/base_tf_config.json'
tf_names = [ TF_loader=TF_loader()
## Geometric TF ##
'Identity',
'FlipUD',
'FlipLR',
'Rotate',
'TranslateX',
'TranslateY',
'ShearX',
'ShearY',
## Color TF (Expect image in the range of [0, 1]) ##
'Contrast',
'Color',
'Brightness',
'Sharpness',
'Posterize',
'Solarize', #=>Image entre [0,1] #Pas opti pour des batch
]
device = torch.device('cuda') #Select device to use device = torch.device('cuda') #Select device to use
@ -48,19 +29,19 @@ if __name__ == "__main__":
}, },
'Inner':{ 'Inner':{
'optim': 'SGD', 'optim': 'SGD',
'lr':1e-2, #1e-2 'lr':1e-2, #1e-2/1e-1 (ResNet)
'momentum':0.9, #0.9 'momentum':0.9, #0.9
'decay':0.0005, #0.0005
'nesterov':False, #False (True: Bad behavior w/ Data_aug)
'scheduler':'cosine', #None, 'cosine', 'multiStep', 'exponential'
} }
} }
#Models #Models
model = LeNet(3,10) model = LeNet(3,10)
#model = ResNet(num_classes=10)
#model = MobileNetV2(num_classes=10)
#model = WideResNet(num_classes=10, wrn_size=32)
#Smart_aug initialisation #Smart_aug initialisation
tf_dict = {k: TF.TF_dict[k] for k in tf_names} tf_dict, tf_ignore_mag =TF_loader.load_TF_dict(tf_config)
model = Higher_model(model) #run_dist_dataugV3 model = Higher_model(model) #run_dist_dataugV3
aug_model = Augmented_model( aug_model = Augmented_model(
Data_augV5(TF_dict=tf_dict, Data_augV5(TF_dict=tf_dict,
@ -68,7 +49,8 @@ if __name__ == "__main__":
mix_dist=0.8, mix_dist=0.8,
fixed_prob=False, fixed_prob=False,
fixed_mag=False, fixed_mag=False,
shared_mag=False), shared_mag=False,
TF_ignore_mag=tf_ignore_mag),
model).to(device) model).to(device)
print("{} on {} for {} epochs - {} inner_it".format(str(aug_model), device_name, epochs, n_inner_iter)) print("{} on {} for {} epochs - {} inner_it".format(str(aug_model), device_name, epochs, n_inner_iter))

66
higher/smart_aug/test_dataug.py
View file

@ -7,58 +7,7 @@ from dataug import *
#from utils import * #from utils import *
from train_utils import * from train_utils import *
# Use available TF (see transformations.py) postfix=''
'''
tf_names = [
## Geometric TF ##
'Identity',
'FlipUD',
'FlipLR',
'Rotate',
'TranslateX',
'TranslateY',
'ShearX',
'ShearY',
#'TranslateXabs',
#'TranslateYabs',
## Color TF (Expect image in the range of [0, 1]) ##
'Contrast',
'Color',
'Brightness',
'Sharpness',
'Posterize',
'Solarize',
#Color TF (Common mag scale)
#'+Contrast',
#'+Color',
#'+Brightness',
#'+Sharpness',
#'-Contrast',
#'-Color',
#'-Brightness',
#'-Sharpness',
#'=Posterize',
#'=Solarize',
## Bad Tranformations ##
# Bad Geometric TF #
#'BShearX',
#'BShearY',
#'BTranslateX-',
#'BTranslateX-',
#'BTranslateY',
#'BTranslateY-',
#'BadContrast',
#'BadBrightness',
#'Random',
#'RandBlend'
]
'''
TF_loader=TF_loader() TF_loader=TF_loader()
device = torch.device('cuda') #Select device to use device = torch.device('cuda') #Select device to use
@ -84,7 +33,7 @@ if __name__ == "__main__":
} }
#Parameters #Parameters
n_inner_iter = 1 n_inner_iter = 1
epochs = 20 epochs = 2
dataug_epoch_start=0 dataug_epoch_start=0
optim_param={ optim_param={
'Meta':{ 'Meta':{
@ -96,7 +45,7 @@ if __name__ == "__main__":
'lr':1e-1, #1e-2/1e-1 (ResNet) 'lr':1e-1, #1e-2/1e-1 (ResNet)
'momentum':0.9, #0.9 'momentum':0.9, #0.9
'decay':0.0005, #0.0005 'decay':0.0005, #0.0005
'nesterov':True, 'nesterov':False, #False (True: Bad behavior w/ Data_aug)
'scheduler':'cosine', #None, 'cosine', 'multiStep', 'exponential' 'scheduler':'cosine', #None, 'cosine', 'multiStep', 'exponential'
} }
} }
@ -161,7 +110,6 @@ if __name__ == "__main__":
if 'aug_model' in tasks: if 'aug_model' in tasks:
tf_config='../config/base_tf_config.json' tf_config='../config/base_tf_config.json'
tf_dict, tf_ignore_mag =TF_loader.load_TF_dict(tf_config) tf_dict, tf_ignore_mag =TF_loader.load_TF_dict(tf_config)
#tf_dict = {k: TF_dict[k] for k in tf_names}
torch.cuda.reset_max_memory_allocated() #reset_peak_stats torch.cuda.reset_max_memory_allocated() #reset_peak_stats
torch.cuda.reset_max_memory_cached() #reset_peak_stats torch.cuda.reset_max_memory_cached() #reset_peak_stats
@ -170,7 +118,7 @@ if __name__ == "__main__":
model = Higher_model(model, model_name) #run_dist_dataugV3 model = Higher_model(model, model_name) #run_dist_dataugV3
aug_model = Augmented_model( aug_model = Augmented_model(
Data_augV5(TF_dict=tf_dict, Data_augV5(TF_dict=tf_dict,
N_TF=3, N_TF=1,
mix_dist=0.5, mix_dist=0.5,
fixed_prob=False, fixed_prob=False,
fixed_mag=False, fixed_mag=False,
@ -184,10 +132,10 @@ if __name__ == "__main__":
inner_it=n_inner_iter, inner_it=n_inner_iter,
dataug_epoch_start=dataug_epoch_start, dataug_epoch_start=dataug_epoch_start,
opt_param=optim_param, opt_param=optim_param,
print_freq=1, print_freq=20,
unsup_loss=1, unsup_loss=1,
hp_opt=False, hp_opt=False,
save_sample_freq=0) save_sample_freq=None)
exec_time=time.perf_counter() - t0 exec_time=time.perf_counter() - t0
max_allocated = torch.cuda.max_memory_allocated()/(1024.0 * 1024.0) max_allocated = torch.cuda.max_memory_allocated()/(1024.0 * 1024.0)
@ -204,7 +152,7 @@ if __name__ == "__main__":
"Param_names": aug_model.TF_names(), "Param_names": aug_model.TF_names(),
"Log": log} "Log": log}
print(str(aug_model),": acc", out["Accuracy"], "in:", out["Time"][0], "+/-", out["Time"][1]) print(str(aug_model),": acc", out["Accuracy"], "in:", out["Time"][0], "+/-", out["Time"][1])
filename = "{}-{} epochs (dataug:{})- {} in_it".format(str(aug_model),epochs,dataug_epoch_start,n_inner_iter) filename = "{}-{} epochs (dataug:{})- {} in_it".format(str(aug_model),epochs,dataug_epoch_start,n_inner_iter)+postfix
with open("../res/log/%s.json" % filename, "w+") as f: with open("../res/log/%s.json" % filename, "w+") as f:
try: try:
json.dump(out, f, indent=True) json.dump(out, f, indent=True)

2
higher/smart_aug/transformations.py
View file

@ -19,7 +19,7 @@ import kornia
import random import random
#TF that don't have use for magnitude parameter. #TF that don't have use for magnitude parameter.
TF_no_mag={'Identity', 'FlipUD', 'FlipLR', 'Random', 'RandBlend', 'identity', 'flipUD', 'flipLR'} TF_no_mag={'Identity', 'FlipUD', 'FlipLR', 'Random', 'RandBlend', 'identity', 'flip'}
#TF which implemetation doesn't allow gradient propagaition. #TF which implemetation doesn't allow gradient propagaition.
TF_no_grad={'Solarize', 'Posterize', '=Solarize', '=Posterize', 'posterize','solarize'} TF_no_grad={'Solarize', 'Posterize', '=Solarize', '=Posterize', 'posterize','solarize'}
#TF for which magnitude should be ignored (Magnitude fixed). #TF for which magnitude should be ignored (Magnitude fixed).