batch_size=4096,
n_epochs=n_epochs,
x=x,
y=y,
x0=x0,
x1=x1,
scale_inputs=True,
#early_stopping = True,
#early_stopping_patience = 10
)
loss_out_path = model_out_path + "/loss"
directory = os.path.dirname(loss_out_path)
if not os.path.exists(loss_out_path):
os.makedirs(loss_out_path)
np.savetxt(loss_out_path + "/train_loss.csv", train_loss, delimiter=",")
np.savetxt(loss_out_path + "/val_loss.csv", val_loss, delimiter=",")
#
model_out_path = model_out_path + '/carl/'
print('all done, saving model to:', model_out_path)
estimator.save(model_out_path, x=x, export_model=True)
# also store patch, if it was provided:
if (args.patch != ""):
patch_out_path = model_out_path + "/patch/"
if not os.path.exists(patch_out_path):
os.makedirs(patch_out_path)
copy2(args.patch, patch_out_path)
TreeName=treename,
randomize=False,
save=True,
correlation=True,
preprocessing=True,
nentries=n,
pathA=p + nominal + ".root",
pathB=p + variation + ".root",
)
logger.info(" Loaded new datasets ")
#######################################
#######################################
# Estimate the likelihood ratio
estimator = RatioEstimator(n_hidden=(10, 10, 10), activation="relu")
estimator.train(
method='carl',
batch_size=1024,
n_epochs=100,
x=x,
y=y,
x0=x0,
x1=x1,
scale_inputs=True,
)
estimator.save('models/' + global_name + '_carl_' + str(n),
x,
metaData,
export_model=True)
########################################
else:
x, y, x0, x1, metaData = loading.loading(
folder='./data/',
plot=True,
var=var,
do=sample,
randomize=False,
save=True,
correlation=True,
preprocessing=True,
nentries=n,
path=p,
)
logger.info(" Loaded new datasets ")
estimator = RatioEstimator(n_hidden=(10, 10, 10), activation="relu")
estimator.train(
method='carl',
batch_size=1024,
n_epochs=100,
x=x,
y=y,
x0=x0,
x1=x1,
scale_inputs=True,
)
estimator.save('models/' + sample + '/' + var + '_carl_' + str(n),
x,
metaData,
export_model=True)