def RunExperiment(trainVal_loader,
test_loader,
epochs,
torchOptim,
lossFn,
net,
results_directory,
name,
device='cpu',
saveAccToo=False,
saveNetDict=True):
losses = train(net,
trainVal_loader,
test_loader,
torchOptim,
lossFn,
0,
epochs,
name,
device=device,
checkForDiv=True,
saveAccToo=saveAccToo,
pathToSave=results_directory,
saveNetDict=saveNetDict)
np.savetxt(results_directory + name + '.txt', np.array(losses), fmt='%s')
def RunExperiment(X_train,
X_val,
X_test,
labels_train,
labels_val,
labels_test,
train_loader,
val_loader,
epochs,
torchOptim,
lossFn,
net,
results_directory,
name,
device='cpu',
saveAccToo=False):
losses = train(net,
train_loader,
val_loader,
torchOptim,
lossFn,
0,
epochs,
name,
device=device,
checkForDiv=True)
np.savetxt(results_directory + name + '.txt', np.array(losses))
def RunExperiment(net, train_loader, train_loader_oversampled, val_loader, \
val_loader_oversampled, Y_tr_np, Y_val_np, epochs, torchOptim, lossFnTr, lossFnVal, \
results_directory, name, lossSigm, device, saveNetDict, scheduler, tol):
losses = train(net, train_loader, train_loader_oversampled, val_loader, \
val_loader_oversampled, Y_tr_np, Y_val_np, torchOptim, lossFnTr, lossFnVal, 0, epochs,\
name, device, lossSigm, checkForDiv = True, pathToSave = results_directory,\
saveNetDict = saveNetDict, scheduler = scheduler, tol = tol)
np.savetxt(results_directory + name + '.txt', np.array(losses), fmt='%s')
def RunExpSavNetAndUsingTest(train_loader, test_loader, epochs, torchOptim, lossFn, net, results_directory, name, device = 'cpu', saveAccToo = False, saveNetDict = True):
losses = train(net, train_loader, test_loader, torchOptim, lossFn, 0, epochs, name, device = device, checkForDiv = True, saveNetDict = saveNetDict, pathToSave = results_directory)
np.savetxt(results_directory + name + '.txt', np.array(losses))