def buildfeaturelib(train_config, model_config):
transform = transforms.Compose(
[transforms.ToTensor(), # 函数接受PIL Image或numpy.ndarray，将其先由HWC转置为CHW格式，再转为float后每个像素除以255.
transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])
mytraindata = myDataset(path=train_config.data_dir, height=model_config.height, width=model_config.width,
autoaugment=model_config.autoaugment, transform=transform)
mytrainloader = torch.utils.data.DataLoader(mytraindata, batch_size=1, shuffle=False)
cuda_gpu = torch.cuda.is_available()
mymodel = builGraph.getModel(model_config.backbone, model_config.num_classes, train_config.gpus,
model_config.modeltype, cuda_gpu=cuda_gpu)
if os.path.exists(train_config.train_dir):
checkpoint = torch.load(train_config.train_dir)
mymodel.load_state_dict(checkpoint['model_state_dict'])
relu_ip1_list = []
logits_list = []
id_list = []
label_list = []
batch_idx=len(mytrainloader)//1
#,batch_x,batch_y,batch_id
mymodel.eval()
with torch.no_grad():
for index,data in enumerate(mytrainloader):
batch_x,batch_y,batch_id=data
if cuda_gpu:
batch_x = batch_x.cuda()
batch_y = batch_y.cuda()
batch_x = batch_x.float()
#batch_x, batch_y = Variable(batch_x), Variable(batch_y)
out,features = mymodel(batch_x)
reluip=features['ip1'].cpu().numpy()
logits=out.cpu().numpy()
batch_y=batch_y.cpu().numpy()
binarvalues=toBinaryString(logits)
for binary in binarvalues:
logits_list.append(binary)
for unit in reluip:
relu_ip1_list.append(unit)
for id in batch_id:
id_list.append(id)
for label in batch_y:
label_list.append(str(label))
if index%10==0:
tool.save(id_list,logits_list,relu_ip1_list,label_list,tofile)
print('Step %d, %.3f%% extracted.' % (index, (index + 1)/ batch_idx * 100))
def getAllData(self, cache=False):
self.initialiseTeams()
print(1)
self.setPlayers()
print(2)
self.setPlayers2kRatings()
print(3)
self.setTeam2kRatings()
print(4)
self.setPlayersMinutesPlayedAndTeamsExpectedQuality()
print(5)
if cache:
save(self.data, 'data.pickle')
if opt.eval:
model.eval()
output_list = []
target_list = []
pbar = tqdm(enumerate(dataset), total=len(dataset))
for i, data in pbar:
if i >= opt.num_test:
break
if "guide_path" in data:
input_path = data["guide_path"][0] # take out the batch
out_path = osp.sep.join(input_path.split(
osp.sep)[-2:])[:-15] # remove _keypoints.json
else:
out_path = str(i)
if i % 5 == 0:
pbar.set_description(out_path)
# process
input = data['A']
target = data['B']
guide = data['guide']
model.set_input(data)
model.test()
output = model.get_output()
# save results
util.save(input, guide, target, output, results_path, out_path, opt)
print('Results saved in %s' % results_path)
guide = data['guide']
model.set_input(data)
model.test()
output = model.get_output()
if (task == 'depth' or task == 'pose'):
# save target/output in a list for evaluation
output_list.append(output)
target_list.append(target)
elif (task == 'texture'):
# save output of textureGAN for evaluation
inpv = torch.cat([input, guide], 1)
textureGAN_output = textureGAN_model(inpv.cuda())
util.save_texture_out(textureGAN_output,
os.path.join(results_path, 'textureGAN'),
i + 1 + (r * dataset_size))
# save results
util.save(input, guide, target, output, results_path,
i + 1 + (r * dataset_size), opt)
print('Results saved in %s' % results_path)
## evaluation
if (task == 'depth'):
print('Evaluating results...')
util.evaluate_depth(output_list, target_list, opt)
elif (task == 'pose'):
print('Evaluating results...')
util.evaluate_pose(output_list, target_list)