AntoineH/smart_augmentation
1
0
Fork 0
mirror of https://github.com/AntoineHX/smart_augmentation.git synced 2025-05-04 12:10:45 +02:00
Code Issues Projects Releases Packages Wiki Activity

modification mineurs

Browse source
This commit is contained in:
Harle, Antoine (Contracteur) 2020-01-13 10:59:32 -05:00
parent 7732060eb1
commit 27c1890f4c
3 changed files with 32 additions and 37 deletions
Download patch file Download diff file Expand all files Collapse all files

3
higher/dataug.py
View file

@ -692,7 +692,8 @@ class Data_augV5(nn.Module): #Optimisation jointe (mag, proba)
else:
return "Data_augV5(Mix%.1f%s-%dTFx%d-%s)" % (self._mix_factor,dist_param, self._nb_tf, self._N_seqTF, mag_param)
class Data_augV6(nn.Module): #Optimisation sequentielle
class Data_augV6(nn.Module): #Optimisation sequentielle #Mauvais resultats
def __init__(self, TF_dict=TF.TF_dict, N_TF=1, mix_dist=0.0, fixed_prob=False, prob_set_size=None, fixed_mag=True, shared_mag=True):
super(Data_augV6, self).__init__()
assert len(TF_dict)>0

64
higher/test_dataug.py
View file

@ -6,21 +6,21 @@ from train_utils import *
tf_names = [
## Geometric TF ##
'Identity',
#'FlipUD',
#'FlipLR',
#'Rotate',
#'TranslateX',
#'TranslateY',
#'ShearX',
#'ShearY',
'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
'Contrast',
'Color',
'Brightness',
'Sharpness',
'Posterize',
'Solarize', #=>Image entre [0,1] #Pas opti pour des batch
#Color TF (Common mag scale)
#'+Contrast',
@ -34,23 +34,18 @@ tf_names = [
#'=Posterize',
#'=Solarize',
#'BRotate',
#'BTranslateX',
#'BTranslateY',
#'BShearX',
#'BShearY',
#'BadTranslateX',
#'BadTranslateX_neg',
#'BadTranslateY',
#'BadTranslateY_neg',
'BShearX',
'BShearY',
'BTranslateX-',
'BTranslateX-',
'BTranslateY',
'BTranslateY-',
#'BadColor',
#'BadSharpness',
#'BadContrast',
#'BadBrightness',
'BadContrast',
'BadBrightness',
'Random',
#'RandBlend'
'RandBlend'
#Non fonctionnel
#'Auto_Contrast', #Pas opti pour des batch (Super lent)
#'Equalize',
@ -71,8 +66,8 @@ if __name__ == "__main__":
#'aug_dataset',
'aug_model'
}
n_inner_iter = 1
epochs = 1
n_inner_iter = 10
epochs = 100
dataug_epoch_start=0
optim_param={
'Meta':{
@ -155,7 +150,7 @@ if __name__ == "__main__":
####
print('-'*9)
times = [x["time"] for x in log]
out = {"Accuracy": max([x["acc"] for x in log]), "Time": (np.mean(times),np.std(times), exec_time), 'Optimizer': optim_param['Inner'], "Device": device_name, "Log": log}
out = {"Accuracy": max([x["acc"] for x in log]), "Time": (np.mean(times),np.std(times), exec_time), 'Optimizer': optim_param['Inner'], "Device": device_name, "Param_names": data_train_aug._TF, "Log": log}
print(str(model),": acc", out["Accuracy"], "in:", out["Time"][0], "+/-", out["Time"][1])
filename = "{}-{}-{} epochs".format(str(data_train_aug),str(model),epochs)
with open("res/log/%s.json" % filename, "w+") as f:
@ -173,8 +168,7 @@ if __name__ == "__main__":
t0 = time.process_time()
tf_dict = {k: TF.TF_dict[k] for k in tf_names}
#aug_model = Augmented_model(Data_augV6(TF_dict=tf_dict, N_TF=1, mix_dist=0.0, fixed_prob=False, prob_set_size=2, fixed_mag=True, shared_mag=True), model).to(device)
aug_model = Augmented_model(Data_augV5(TF_dict=tf_dict, N_TF=1, mix_dist=0.5, fixed_prob=False, fixed_mag=True, shared_mag=True), model).to(device)
aug_model = Augmented_model(Data_augV5(TF_dict=tf_dict, N_TF=2, mix_dist=0.0, fixed_prob=False, fixed_mag=False, shared_mag=False), model).to(device)
#aug_model = Augmented_model(RandAug(TF_dict=tf_dict, N_TF=2), model).to(device)
print("{} on {} for {} epochs - {} inner_it".format(str(aug_model), device_name, epochs, n_inner_iter))
@ -183,7 +177,7 @@ if __name__ == "__main__":
inner_it=n_inner_iter,
dataug_epoch_start=dataug_epoch_start,
opt_param=optim_param,
print_freq=1,
print_freq=10,
KLdiv=True,
loss_patience=None)
@ -191,14 +185,14 @@ if __name__ == "__main__":
####
print('-'*9)
times = [x["time"] for x in log]
out = {"Accuracy": max([x["acc"] for x in log]), "Time": (np.mean(times),np.std(times), exec_time), 'Optimizer': optim_param, "Device": device_name, "Log": log}
out = {"Accuracy": max([x["acc"] for x in log]), "Time": (np.mean(times),np.std(times), exec_time), 'Optimizer': optim_param, "Device": device_name, "Param_names": aug_model.TF_names(), "Log": log}
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)
with open("res/log/%s.json" % filename, "w+") as f:
json.dump(out, f, indent=True)
print('Log :\"',f.name, '\" saved !')
plot_resV2(log, fig_name="res/"+filename, param_names=tf_names)
plot_resV2(log, fig_name="res/"+filename, param_names=aug_model.TF_names())
print('Execution Time : %.00f '%(exec_time))
print('-'*9)