def validation(discriminator, send_stats=False, epoch=0):
print('Validating model on {0} examples. '.format(
len(validation_dataset)))
discriminator_ = discriminator.eval()
with torch.no_grad():
pred_logits_list = []
labels_list = []
for (inp, labels, imgs_names) in tqdm(validation_loader):
inp = Variable(inp.float(), requires_grad=False)
labels = Variable(labels.long(), requires_grad=False)
if hparams.dim3:
inp = inp.view(-1, 1, 640, 64)
inp = torch.cat([inp] * 3, dim=1)
inp = inp.to(hparams.gpu_device)
labels = labels.to(hparams.gpu_device)
pred_logits = discriminator_(inp)
pred_logits_list.append(pred_logits)
labels_list.append(labels)
pred_logits = torch.cat(pred_logits_list, dim=0)
labels = torch.cat(labels_list, dim=0)
val_loss = adversarial_loss(pred_logits, labels)
return accuracy_metrics(
labels.long(), pred_logits
), val_loss #, plot_auc='train_val_'+str(epoch+1), plot_path=hparams.result_dir+'train_val_{}_'.format(epoch)), val_loss
def validation(discriminator, send_stats=False, epoch=0):
print('Validating model on {0} examples. '.format(
len(validation_dataset)))
discriminator_ = discriminator.eval()
with torch.no_grad():
pred_logits_list = []
labels_list = []
for (img, labels, imgs_names) in tqdm(validation_loader):
img = Variable(img.float(), requires_grad=False)
labels = Variable(labels.float(), requires_grad=False)
img_ = img.to(hparams.gpu_device)
labels = labels.to(hparams.gpu_device)
pred_logits = discriminator_(img_)
pred_logits_list.append(pred_logits)
labels_list.append(labels)
pred_logits = torch.cat(pred_logits_list, dim=0)
labels = torch.cat(labels_list, dim=0)
val_loss = adversarial_loss(pred_logits, labels)
return accuracy_metrics(labels.long(), pred_logits), val_loss
def test(model_paths,
data=(hparams.valid_csv, hparams.valid_dir),
plot_auc='valid',
plot_path=hparams.result_dir + 'valid',
best_thresh=None,
pred_csv=None):
test_dataset = ChestData(
data_csv=data[0],
data_dir=data[1],
augment=hparams.TTA,
transform=transforms.Compose([
transforms.Resize(hparams.image_shape),
transforms.ToTensor(),
# transforms.Normalize((0.5027, 0.5027, 0.5027), (0.2915, 0.2915, 0.2915))
]))
test_loader = DataLoader(test_dataset,
batch_size=hparams.batch_size,
shuffle=False,
num_workers=4)
discriminators = [
Discriminator().to(hparams.gpu_device) for _ in model_paths
]
if hparams.cuda:
discriminators = [
nn.DataParallel(discriminators[i], device_ids=hparams.device_ids)
for i in range(len(model_paths))
]
checkpoints = [
torch.load(model_path, map_location=hparams.gpu_device)
for model_path in model_paths
]
for i in range(len(model_paths)):
discriminators[i].load_state_dict(
checkpoints[i]['discriminator_state_dict'])
def put_eval(model):
model = model.eval()
model.training = hparams.eval_dp_on
return model
discriminators = [
put_eval(discriminator) for discriminator in discriminators
]
# print('Model loaded')
Tensor = torch.cuda.FloatTensor if hparams.cuda else torch.FloatTensor
print('Testing model on {0} examples. '.format(len(test_dataset)))
with torch.no_grad():
pred_logits = torch.zeros((len(test_dataset), hparams.num_classes))
if hparams.cuda:
pred_logits = pred_logits.to(hparams.gpu_device)
for _ in range(hparams.repeat_infer):
labels_list = []
img_names_list = []
pred_logits_list = []
for (img, labels, img_names) in tqdm(test_loader):
img = Variable(img.float(), requires_grad=False)
labels = Variable(labels.float(), requires_grad=False)
if hparams.cuda:
img_ = img.to(hparams.gpu_device)
labels = labels.to(hparams.gpu_device)
pred_logits_ = discriminators[0](img_)
pred_logits_ = pred_logits_ * 0
for discriminator in discriminators:
pred_logits_ += discriminator(img_)
pred_logits_ = 1.0 * pred_logits_ / len(model_paths)
pred_logits_list.append(pred_logits_)
labels_list.append(labels)
img_names_list += list(img_names)
pred_logits += torch.cat(pred_logits_list, dim=0)
labels = torch.cat(labels_list, dim=0)
pred_logits = 1.0 * pred_logits / hparams.repeat_infer
_, pred_labels = torch.max(pred_logits, axis=1)
f1, acc, conf_mat = accuracy_metrics(labels, pred_labels)
print('== Test on -- '+str(model_paths)+' == \n f1 - {0:.4f}, acc - {1:.4f}'\
.format(f1, acc))
return f1
def test(model_path,
data=(hparams.valid_csv, hparams.dev_file),
plot_auc='valid',
plot_path=hparams.result_dir + 'valid',
best_thresh=None):
test_dataset = AudioData(data_csv=data[0],
data_file=data[1],
ds_type='valid',
augment=True,
transform=transforms.Compose([
transforms.ToTensor(),
]))
test_loader = DataLoader(test_dataset,
batch_size=hparams.batch_size,
shuffle=True,
num_workers=2)
discriminator = Discriminator().to(hparams.gpu_device)
if hparams.cuda:
discriminator = nn.DataParallel(discriminator,
device_ids=hparams.device_ids)
checkpoint = torch.load(model_path, map_location=hparams.gpu_device)
discriminator.load_state_dict(checkpoint['discriminator_state_dict'])
discriminator = discriminator.eval()
# print('Model loaded')
Tensor = torch.cuda.FloatTensor if hparams.cuda else torch.FloatTensor
print('Testing model on {0} examples. '.format(len(test_dataset)))
with torch.no_grad():
pred_logits_list = []
labels_list = []
img_names_list = []
# for _ in range(hparams.repeat_infer):
for (inp, labels, img_names) in tqdm(test_loader):
inp = Variable(inp.float(), requires_grad=False)
labels = Variable(labels.long(), requires_grad=False)
inp = inp.to(hparams.gpu_device)
labels = labels.to(hparams.gpu_device)
if hparams.dim3:
inp = inp.view(-1, 1, 640, 64)
inp = torch.cat([inp] * 3, dim=1)
pred_logits = discriminator(inp)
pred_logits_list.append(pred_logits)
labels_list.append(labels)
img_names_list.append(img_names)
pred_logits = torch.cat(pred_logits_list, dim=0)
labels = torch.cat(labels_list, dim=0)
auc, f1, acc, conf_mat = accuracy_metrics(labels,
pred_logits,
plot_auc=plot_auc,
plot_path=plot_path,
best_thresh=best_thresh)
fig = plot_cf(conf_mat)
plt.savefig(hparams.result_dir + 'test_conf_mat.png')
res = ' -- avg_acc - {0:.4f}'.format(acc['avg'])
for it in range(10):
res += ', acc_{}'.format(
hparams.id_to_class[it]) + ' - {0:.4f}'.format(acc[it])
print('== Test on -- ' + model_path + res)
# print('== Test on -- '+model_path+' == \n\
# auc_{0} - {10:.4f}, auc_{1} - {11:.4f}, auc_{2} - {12:.4f}, auc_{3} - {13:.4f}, auc_{4} - {14:.4f}, auc_{5} - {15:.4f}, auc_{6} - {16:.4f}, auc_{7} - {17:.4f}, auc_{8} - {18:.4f}, auc_{9} - {19:.4f}, auc_micro - {20:.4f}, auc_macro - {21:.4f},\n\
# acc_{0} - {22:.4f}, acc_{1} - {23:.4f}, acc_{2} - {24:.4f}, acc_{3} - {25:.4f}, acc_{4} - {26:.4f}, acc_{5} - {27:.4f}, acc_{6} - {28:.4f}, acc_{7} - {29:.4f}, acc_{8} - {30:.4f}, acc_{9} - {31:.4f}, acc_avg - {32:.4f},\n\
# f1_{0} - {33:.4f}, f1_{1} - {34:.4f}, f1_{2} - {35:.4f}, f1_{3} - {36:.4f}, f1_{4} - {37:.4f}, f1_{5} - {38:.4f}, f1_{6} - {39:.4f}, f1_{7} - {40:.4f}, f1_{8} - {41:.4f}, f1_{9} - {42:.4f}, f1_micro - {42:.4f}, f1_macro - {43:.4f}, =='.\
# format([hparams.id_to_class[it] for it in range(10)]+[auc[it] for it in range(10)]+[auc['micro'], auc['macro']]+[acc[it] for it in range(10)]+[acc['avg']]+[f1[it] for it in range(10)]+[f1['micro'], f1['macro']]))
return acc['avg']
def train(resume=False):
writer = SummaryWriter('../runs/' + hparams.exp_name)
for k in hparams.__dict__.keys():
writer.add_text(str(k), str(hparams.__dict__[k]))
train_dataset = AudioData(
data_csv=hparams.train_csv,
data_file=hparams.dev_file,
ds_type='train', # augment=False,
transform=transforms.Compose([
transforms.ToTensor(),
]))
validation_dataset = AudioData(data_csv=hparams.valid_csv,
data_file=hparams.dev_file,
ds_type='valid',
augment=False,
transform=transforms.Compose([
transforms.ToTensor(),
]))
# train_sampler = WeightedRandomSampler()
train_loader = DataLoader(train_dataset,
batch_size=hparams.batch_size,
shuffle=True,
num_workers=2)
validation_loader = DataLoader(validation_dataset,
batch_size=hparams.batch_size,
shuffle=True,
num_workers=2)
print('loaded train data of length : {}'.format(len(train_dataset)))
adversarial_loss = torch.nn.CrossEntropyLoss().to(hparams.gpu_device)
discriminator = Discriminator().to(hparams.gpu_device)
if hparams.cuda:
discriminator = nn.DataParallel(discriminator,
device_ids=hparams.device_ids)
params_count = 0
for param in discriminator.parameters():
params_count += np.prod(param.size())
print('Model has {0} trainable parameters'.format(params_count))
if not hparams.pretrained:
discriminator.apply(weights_init_normal)
optimizer_D = torch.optim.Adam(discriminator.parameters(),
lr=hparams.learning_rate)
scheduler_D = ReduceLROnPlateau(optimizer_D,
mode='min',
factor=0.3,
patience=4,
verbose=True,
cooldown=0)
Tensor = torch.cuda.FloatTensor if hparams.cuda else torch.FloatTensor
def validation(discriminator, send_stats=False, epoch=0):
print('Validating model on {0} examples. '.format(
len(validation_dataset)))
discriminator_ = discriminator.eval()
with torch.no_grad():
pred_logits_list = []
labels_list = []
for (inp, labels, imgs_names) in tqdm(validation_loader):
inp = Variable(inp.float(), requires_grad=False)
labels = Variable(labels.long(), requires_grad=False)
if hparams.dim3:
inp = inp.view(-1, 1, 640, 64)
inp = torch.cat([inp] * 3, dim=1)
inp = inp.to(hparams.gpu_device)
labels = labels.to(hparams.gpu_device)
pred_logits = discriminator_(inp)
pred_logits_list.append(pred_logits)
labels_list.append(labels)
pred_logits = torch.cat(pred_logits_list, dim=0)
labels = torch.cat(labels_list, dim=0)
val_loss = adversarial_loss(pred_logits, labels)
return accuracy_metrics(
labels.long(), pred_logits
), val_loss #, plot_auc='train_val_'+str(epoch+1), plot_path=hparams.result_dir+'train_val_{}_'.format(epoch)), val_loss
print('Starting training.. (log saved in:{})'.format(hparams.exp_name))
start_time = time.time()
best_valid_acc = 0
# print(model)
for epoch in range(hparams.num_epochs):
train_logits = []
train_labels = []
for batch, (inp, labels, imgs_name) in enumerate(tqdm(train_loader)):
inp = Variable(inp.float(), requires_grad=False)
labels = Variable(labels.long(), requires_grad=False)
inp = inp.to(hparams.gpu_device)
labels = labels.to(hparams.gpu_device)
if hparams.dim3:
inp = inp.view(-1, 1, 640, 64)
inp = torch.cat([inp] * 3, dim=1)
# ---------------------
# Train Discriminator
# ---------------------
optimizer_D.zero_grad()
pred_logits = discriminator(inp)
train_logits.append(pred_logits)
train_labels.append(labels)
d_loss = adversarial_loss(pred_logits, labels)
d_loss.backward()
optimizer_D.step()
writer.add_scalar('d_loss',
d_loss.item(),
global_step=batch + epoch * len(train_loader))
# if batch % hparams.print_interval == 0:
# pred_labels = (pred_logits >= hparams.thresh)
# pred_labels = pred_labels.float()
# auc, f1, acc, _, _ = accuracy_metrics(pred_labels, labels.long(), pred_logits)
# print('[Epoch - {0:.1f}, batch - {1:.3f}, d_loss - {2:.6f}, acc - {3:.4f}, f1 - {4:.5f}, auc - {5:.4f}]'.\
# format(1.0*epoch, 100.0*batch/len(train_loader), d_loss.item(), acc['avg'], f1[hparams.avg_mode], auc[hparams.avg_mode]))
(val_auc, val_f1, val_acc,
val_conf_mat), val_loss = validation(discriminator, epoch=epoch)
train_logits = torch.cat(train_logits, dim=0)
train_labels = torch.cat(train_labels, dim=0)
train_auc, train_f1, train_acc, train_conf_mat = accuracy_metrics(
train_labels.long(), train_logits)
fig = plot_cf(val_conf_mat)
writer.add_figure('val_conf', fig, global_step=epoch)
plt.close(fig)
for lbl in range(hparams.num_classes):
writer.add_scalar('val_f1_{}'.format(hparams.id_to_class[lbl]),
val_f1[lbl],
global_step=epoch)
writer.add_scalar('val_auc_{}'.format(hparams.id_to_class[lbl]),
val_auc[lbl],
global_step=epoch)
writer.add_scalar('val_acc_{}'.format(hparams.id_to_class[lbl]),
val_acc[lbl],
global_step=epoch)
writer.add_scalar('val_f1_{}'.format('micro'),
val_f1['micro'],
global_step=epoch)
writer.add_scalar('val_auc_{}'.format('micro'),
val_auc['micro'],
global_step=epoch)
writer.add_scalar('val_f1_{}'.format('macro'),
val_f1['macro'],
global_step=epoch)
writer.add_scalar('val_auc_{}'.format('macro'),
val_auc['macro'],
global_step=epoch)
writer.add_scalar('val_loss', val_loss, global_step=epoch)
writer.add_scalar('val_f1',
val_f1[hparams.avg_mode],
global_step=epoch)
writer.add_scalar('val_auc',
val_auc[hparams.avg_mode],
global_step=epoch)
writer.add_scalar('val_acc', val_acc['avg'], global_step=epoch)
scheduler_D.step(val_loss)
writer.add_scalar('learning_rate',
optimizer_D.param_groups[0]['lr'],
global_step=epoch)
# torch.save({
# 'epoch': epoch,
# 'discriminator_state_dict': discriminator.state_dict(),
# 'optimizer_D_state_dict': optimizer_D.state_dict(),
# }, hparams.model+'.'+str(epoch))
if best_valid_acc <= val_acc['avg']:
best_valid_acc = val_acc['avg']
fig = plot_cf(val_conf_mat)
writer.add_figure('best_val_conf', fig, global_step=epoch)
plt.close(fig)
torch.save(
{
'epoch': epoch,
'discriminator_state_dict': discriminator.state_dict(),
'optimizer_D_state_dict': optimizer_D.state_dict(),
}, hparams.model + '.best')
print('best model on validation set saved.')
print('[Epoch - {0:.1f} ---> train_acc - {1:.4f}, current_lr - {2:.6f}, val_loss - {3:.4f}, best_val_acc - {4:.4f}, val_acc - {5:.4f}, val_f1 - {6:.4f}] - time - {7:.1f}'\
.format(1.0*epoch, train_acc['avg'], optimizer_D.param_groups[0]['lr'], val_loss, best_valid_acc, val_acc['avg'], val_f1[hparams.avg_mode], time.time()-start_time))
start_time = time.time()
def test(model_paths, data=(hparams.valid_csv, hparams.valid_dir), plot_auc='valid', plot_path=hparams.result_dir+'valid', best_thresh=None, pred_csv=None):
test_dataset = ChestData(data_csv=data[0], data_dir=data[1], augment=hparams.TTA,
transform=transforms.Compose([
transforms.Resize(hparams.image_shape),
transforms.ToTensor(),
# transforms.Normalize((0.5027, 0.5027, 0.5027), (0.2915, 0.2915, 0.2915))
]))
test_loader = DataLoader(test_dataset, batch_size=hparams.batch_size,
shuffle=False, num_workers=4)
discriminators = [Discriminator().to(hparams.gpu_device) for _ in model_paths]
if hparams.cuda:
discriminators = [nn.DataParallel(discriminators[i], device_ids=hparams.device_ids) for i in range(len(model_paths))]
checkpoints = [torch.load(model_path, map_location=hparams.gpu_device) for model_path in model_paths]
for i in range(len(model_paths)):
# print(checkpoints[i])
discriminators[i].load_state_dict(checkpoints[i]['discriminator_state_dict'])
def put_eval(model):
model = model.eval()
# model.training = hparams.eval_dp_on
return model
discriminators = [put_eval(discriminator) for discriminator in discriminators]
# print('Model loaded')
Tensor = torch.cuda.FloatTensor if hparams.cuda else torch.FloatTensor
print('Testing model on {0} examples. '.format(len(test_dataset)))
with torch.no_grad():
pred_logits = torch.zeros((len(test_dataset), hparams.num_classes))
if hparams.cuda:
pred_logits = pred_logits.to(hparams.gpu_device)
for _ in range(hparams.repeat_infer):
labels_list = []
img_names_list = []
pred_logits_list = []
for (img, labels, img_names) in tqdm(test_loader):
img = Variable(img.float(), requires_grad=False)
labels = Variable(labels.float(), requires_grad=False)
if hparams.cuda:
img_ = img.to(hparams.gpu_device)
labels = labels.to(hparams.gpu_device)
pred_logits_ = discriminators[0](img_)
pred_logits_ = pred_logits_*0
for discriminator in discriminators:
pred_logits_ += discriminator(img_)
pred_logits_ = 1.0*pred_logits_/len(model_paths)
pred_logits_list.append(pred_logits_)
labels_list.append(labels)
img_names_list += list(img_names)
pred_logits += torch.cat(pred_logits_list, dim=0)
labels = torch.cat(labels_list, dim=0)
pred_logits = 1.0*pred_logits/hparams.repeat_infer
print(plot_path)
# pred_logits = aggr_preds(img_names_list, pred_logits, 'min')
# labels = aggr_preds(img_names_list, labels, 'min')
auc, f1, acc, conf_mat, best_thresh = accuracy_metrics(labels, pred_logits, plot_auc=plot_auc, plot_path=plot_path, best_thresh=best_thresh)
if hparams.cuda:
pred_logits = pred_logits.cpu()
pred_logits = pred_logits.numpy()
print(best_thresh)
pred_labels = 1*(pred_logits > np.array(best_thresh))
if pred_csv:
data = {'Path': img_names_list}
for lbl in range(14):
data[hparams.id_to_class[lbl]] = pred_labels[:,lbl]
df = pd.DataFrame(data)
df.to_csv('../results/predictions_{}.csv'.format(pred_csv), index=False)
print('predictions saved to "../results/predictions_{}.csv"'.format(pred_csv))
print('== Test on -- '+str(model_paths)+' == \n\
auc_{0} - {5:.4f}, auc_{1} - {6:.4f}, auc_{2} - {7:.4f}, auc_{3} - {8:.4f}, auc_{4} - {9:.4f}, auc_micro - {10:.4f}, auc_macro - {11:.4f},\n\
acc_{0} - {12:.4f}, acc_{1} - {13:.4f}, acc_{2} - {14:.4f}, acc_{3} - {15:.4f}, acc_{4} - {16:.4f}, acc_avg - {17:.4f},\n\
f1_{0} - {18:.4f}, f1_{1} - {19:.4f}, f1_{2} - {20:.4f}, f1_{3} - {21:.4f}, f1_{4} - {22:.4f}, f1_micro - {23:.4f}, f1_macro - {24:.4f},\n\
thresh_{0} - {25:4f}, thresh_{1} - {26:4f}, thresh_{2} - {27:4f}, thresh_{3} - {28:4f}, thresh_{4} - {29:4f} =='.\
format(hparams.id_to_class[0], hparams.id_to_class[1], hparams.id_to_class[2], hparams.id_to_class[3], hparams.id_to_class[4], auc[0], auc[1], auc[2], auc[3], auc[4], auc['micro'], auc['macro'], acc[0], acc[1], acc[2], acc[3], acc[4], acc['avg'],
f1[0], f1[1], f1[2], f1[3], f1[4], f1['micro'], f1['macro'], best_thresh[0], best_thresh[1], best_thresh[2], best_thresh[3], best_thresh[4]))
return auc['micro']
def test(model_path=hparams.model, send_stats=False):
test_dataset = APTOSData(csv_file=hparams.train_csv,
root_dir=hparams.train_dir,
split=0.95,
test_split=0.95,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))
]),
ds_type='test', file_format='.png')
extra_test_dataset = APTOSData(csv_file=hparams.dg_test_csv,
root_dir=hparams.dg_test_dir,
split=0,
test_split=0.0,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))
]),
ds_type='test', file_format='.jpg')
test_loader = DataLoader(test_dataset, batch_size=64,
shuffle=False, num_workers=2)
extra_test_loader = DataLoader(extra_test_dataset, batch_size=64,
shuffle=False, num_workers=2)
test_loaders = [test_loader, extra_test_loader]
# print('loaded test data of length :'+str(len(test_loader)))
if hparams.cuda:
model = Rater(hparams.image_shape, hparams.num_classes, pretrained=False).cuda(hparams.gpu_device)
checkpoint = torch.load(model_path)
model.load_state_dict(checkpoint['model_state_dict'])
else:
model = Rater(hparams.image_shape, hparams.num_classes, pretrained=False)
checkpoint = torch.load(model_path, map_location='cpu')
model.load_state_dict(checkpoint['model_state_dict'])
# print('Model loaded')
Tensor = torch.cuda.FloatTensor if hparams.cuda else torch.FloatTensor
rounder = OptimizedRounder()
print('Testing model on {0} examples. '.format(len(test_loader)))
with open('submission.csv', 'w') as f:
writer = csv.writer(f)
writer.writerow(['id_code', 'diagnosis'])
with torch.no_grad():
pred_labels_list = []
labels_list = []
img_names_list = []
for loader in tqdm(test_loaders):
for i, (img, labels, img_names) in enumerate(loader):
img = Variable(img.type(Tensor), requires_grad=False)
labels = Variable(labels.float(), requires_grad=False)
if hparams.cuda:
img = img.cuda(hparams.gpu_device)
labels = labels.cuda(hparams.gpu_device)
pred_logits, _ = model(img)
pred_labels = torch.tensor(rounder.predict(pred_logits.view(-1), hparams.coefficients))
img_names_list += list(img_names)
pred_labels_list.append(pred_labels.view(-1))
labels_list.append(labels.view(-1))
pred_labels = torch.cat(pred_labels_list, dim=0)
labels = torch.cat(labels_list, dim=0)
for i in range(len(img_names_list)):
writer.writerow([img_names_list[i], pred_labels[i].item()])
if send_stats:
kappa, precision, recall, f1, accuracy, precision_list, recall_list, f1_list, accuracy_list = accuracy_metrics(pred_labels, labels, True)
kappa, precision, recall, f1, accuracy = accuracy_metrics(pred_labels, labels, False)
print('== Test on -- '+model_path+' == kappa - {0:.4f}, precision - {1:.4f}, recall - {2:.4f}, f1 - {3:.4f}, accuracy - {4:.4f} =='.format(kappa, precision, recall, f1, accuracy))
return kappa
