loss.backward()
net.adam.step()
err = loss.cpu().data.numpy()[0]
return err
net = ClassifierGenerator(FEATURES=2, CLASSES=4, NETSIZE=384).cuda()
difficulty_level = 1.0
errs = []
err = 0
err_count = 0
for i in range(40000):
err += trainingStep(net, 400, min_difficulty = difficulty_level, max_difficulty = difficulty_level, feature_variation = False, class_variation = False)
err_count += 1
if err_count >= 50:
err = err/err_count
errs.append(err)
f = open("training_curves/training2-base.txt","a")
f.write("%d %.6g %.6g\n" % (i, err, difficulty_level))
f.close()
err = 0
err_count = 0
torch.save(net.state_dict(),open("models/classifier-generator-2-4-base.pth","wb"))
difficulty_level = 0.0125
errs = []
err = 0
err_count = 0
for i in range(100000):
err += trainingStep(net,
100,
min_difficulty=difficulty_level * 0.5,
max_difficulty=difficulty_level * 1.5)
err_count += 1
if i % 10000 == 5000:
torch.save(
net.state_dict(),
open("ckpt/classifier-generator-128-16-ckpt%d.pth" % i, "wb"))
if err_count >= 50:
err = err / err_count
errs.append(err)
methods = [lambda: NetworkSKL(net)]
results1 = compareMethodsOnSet(methods,
echocardio['x'],
echocardio['y'].astype(np.int32),
samples=200)
auc1 = results1[0][1]
results2 = compareMethodsOnSet(methods,
bloodtransfusion['x'],
bloodtransfusion['y'].astype(np.int32),
for file in glob.glob("data/*.npz"):
data = np.load(file)
if np.unique(data['y']).shape[0] <= 16:
data_names.append(file[5:-4])
data_x.append(data['x'].copy())
data_y.append(data['y'].copy().astype(np.int32))
for didx in range(len(data_names)):
net = ClassifierGenerator(FEATURES=128, CLASSES=16, NETSIZE=384).cuda()
net.load_state_dict(torch.load("models/classifier-generator-128-16.pth"))
tdx = []
tdy = []
for didx2 in range(len(data_names)):
if didx2 != didx:
if data_x[didx2].shape[0] >= 120:
tdx.append(data_x[didx2])
tdy.append(data_y[didx2])
for i in range(20):
err = trainingStep(net, 100, 20, tdx, tdy)
f = open("training_curves/finetuning-%s.txt" % data_names[didx], "a")
f.write("%d %.6g\n" % (i, err))
f.close()
torch.save(
net.state_dict(),
open("models/classifier-generator-128-16-%s.pth" % data_names[didx],
"wb"))
errs = []
err = 0
err_count = 0
for i in range(40000):
err += trainingStep(net, 20+np.random.randint(380), min_difficulty = 0.25, max_difficulty = 4.0, feature_variation = False, class_variation = False)
err_count += 1
if err_count >= 50:
err = err/err_count
errs.append(err)
#methods = [lambda: NetworkSKL(net)]
#results1 = compareMethodsOnSet(methods, echocardio['x'], echocardio['y'].astype(np.int32), samples=200)
#auc1 = results1[0][1]
#results2 = compareMethodsOnSet(methods, bloodtransfusion['x'], bloodtransfusion['y'].astype(np.int32), samples=200)
#auc2 = results2[0][1]
#results3 = compareMethodsOnSet(methods, autism['x'], autism['y'].astype(np.int32), samples=200)
#auc3 = results3[0][1]
f = open("training2-general.txt","a")
f.write("%d %.6g\n" % (i, err))
f.close()
err = 0
err_count = 0
torch.save(net.state_dict(),open("classifier-generator-2-4-general.pth","wb"))