class Logger(object):
def __init__(self, log_dir):
"""Create a summary writer logging to log_dir."""
self.writer = SummaryWriter(log_dir=log_dir)
def scalar_summary(self, tag, value, step):
"""Log a scalar variable."""
self.writer.add_summary(tag, value, step)
def image_summary(self, tag, images, step):
"""Log a list of images."""
self.writer.add_image(tag, images, step)
def graph_summary(self, model, dummy_input):
self.writer.add_graph(model, (dummy_input, ))
class Logger(object):
def __init__(self, log_dir, step=0):
"""Create a summary writer logging to log_dir."""
log_dir = os.path.join('log', log_dir)
print('logging at: ', log_dir)
self.writer = SummaryWriter(log_dir)
self.step = step
def incstep(self):
self.step += 1
def scalar_summary(self, data):
"""Log a scalar variable."""
for tag, value in data.items():
self.writer.add_scalar(tag, value, self.step)
def image_summary(self, tag, images, step):
"""Log a list of images."""
img_summaries = []
for i, img in enumerate(images):
# Write the image to a string
try:
s = StringIO()
except:
s = BytesIO()
scipy.misc.toimage(img).save(s, format="png")
# Create an Image object
img_sum = tf.Summary.Image(encoded_image_string=s.getvalue(),
height=img.shape[0],
width=img.shape[1])
# Create a Summary value
img_summaries.append(
tf.Summary.Value(tag='%s/%d' % (tag, i), image=img_sum))
# Create and write Summary
summary = tf.Summary(value=img_summaries)
self.writer.add_summary(summary, step)
def hist_summary(self, tag, values):
"""Log a histogram of the tensor of values."""
self.writer.add_histogram(tag, values, self.step, bins="auto")
class Logger(object):
def __init__(self, log_dir):
"""Create a summary writer logging to log_dir."""
self.writer = SummaryWriter(log_dir=log_dir)
def add_scalars(self, tag, value_dict, step):
"""Log a scalar variable."""
self.writer.add_scalars(tag, value_dict, step)
def image_summary(self, tag, images, step):
"""Log a list of images."""
img_summaries = []
for i, img in enumerate(images):
# Write the image to a string
try:
s = StringIO()
except:
s = BytesIO()
scipy.misc.toimage(img).save(s, format="png")
# Create an Image object
img_sum = tf.Summary.Image(encoded_image_string=s.getvalue(),
height=img.shape[0],
width=img.shape[1])
# Create a Summary value
img_summaries.append(
tf.Summary.Value(tag='%s/%d' % (tag, i), image=img_sum))
# Create and write Summary
summary = tf.Summary(value=img_summaries)
self.writer.add_summary(summary, step)
def add_hist(self, tag, values, step, bins=1000):
"""Log a histogram of the tensor of values."""
# self.writer.add_histogram('hist', array, iteration)
pass
class Logger(object):
def __init__(self, log_dir):
"""Create a summary writer logging to log_dir."""
log_dir = os.path.join('log',log_dir)
if not os.path.exists(log_dir):
print('log dir', log_dir)
#self.writer = tf.compat.v1.summary.FileWriter(log_dir)
self.writer = SummaryWriter(log_dir)
else:
print('This training session name already exists. Please use a different name or delete it')
self.step = 0
def incstep(self): self.step += 1
def scalar_summary(self, data):
"""Log a scalar variable."""
for tag, value in data.items():
self.writer.add_scalar(tag,value,self.step)
#summary = tf.compat.v1.Summary(value=[tf.compat.v1.Summary.Value(tag=tag, simple_value=value)])
#self.writer.add_summary(summary, self.step)
def image_summary(self, tag, images, step):
"""Log a list of images."""
img_summaries = []
for i, img in enumerate(images):
# Write the image to a string
try:
s = StringIO()
except:
s = BytesIO()
scipy.misc.toimage(img).save(s, format="png")
# Create an Image object
img_sum = tf.Summary.Image(encoded_image_string=s.getvalue(),
height=img.shape[0],
width=img.shape[1])
# Create a Summary value
img_summaries.append(tf.Summary.Value(tag='%s/%d' % (tag, i), image=img_sum))
# Create and write Summary
summary = tf.Summary(value=img_summaries)
self.writer.add_summary(summary, step)
def hist_summary(self, tag, values, bins=5):
"""Log a histogram of the tensor of values."""
'''
# Create a histogram using numpy
counts, bin_edges = np.histogram(values, bins=bins)
# Fill the fields of the histogram proto
hist = tf.compat.v1.HistogramProto()
hist.min = float(np.min(values))
hist.max = float(np.max(values))
hist.num = int(np.prod(values.shape))
hist.sum = float(np.sum(values))
hist.sum_squares = float(np.sum(values**2))
# Drop the start of the first bin
bin_edges = bin_edges[1:]
# Add bin edges and counts
for edge in bin_edges:
hist.bucket_limit.append(edge)
for c in counts:
hist.bucket.append(c)
'''
# Create and write Summary
print('hello')
self.writer.add_histogram(tag,values.clone().cpu().data.numpy(),self.step)
print('hello')
'''
class Visualizer:
def __init__(self, opt):
self.opt = opt
self.tf_log = opt.isTrain and opt.tf_log
self.tensorboard_log = opt.tensorboard_log
self.win_size = opt.display_winsize
self.name = opt.name
if self.tensorboard_log:
if self.opt.isTrain:
self.log_dir = os.path.join(opt.checkpoints_dir, opt.name, "logs")
if not os.path.exists(self.log_dir):
os.makedirs(self.log_dir)
self.writer = SummaryWriter(log_dir=self.log_dir)
else:
print("hi :)")
self.log_dir = os.path.join(opt.checkpoints_dir, opt.name, opt.results_dir)
if not os.path.exists(self.log_dir):
os.makedirs(self.log_dir)
if opt.isTrain:
self.log_name = os.path.join(opt.checkpoints_dir, opt.name, "loss_log.txt")
with open(self.log_name, "a") as log_file:
now = time.strftime("%c")
log_file.write("================ Training Loss (%s) ================\n" % now)
# |visuals|: dictionary of images to display or save
def display_current_results(self, visuals, epoch, step):
all_tensor = []
if self.tensorboard_log:
for key, tensor in visuals.items():
all_tensor.append((tensor.data.cpu() + 1) / 2)
output = torch.cat(all_tensor, 0)
img_grid = vutils.make_grid(output, nrow=self.opt.batchSize, padding=0, normalize=False)
if self.opt.isTrain:
self.writer.add_image("Face_SPADE/training_samples", img_grid, step)
else:
vutils.save_image(
output,
os.path.join(self.log_dir, str(step) + ".png"),
nrow=self.opt.batchSize,
padding=0,
normalize=False,
)
# errors: dictionary of error labels and values
def plot_current_errors(self, errors, step):
if self.tf_log:
for tag, value in errors.items():
value = value.mean().float()
summary = self.tf.Summary(value=[self.tf.Summary.Value(tag=tag, simple_value=value)])
self.writer.add_summary(summary, step)
if self.tensorboard_log:
self.writer.add_scalar("Loss/GAN_Feat", errors["GAN_Feat"].mean().float(), step)
self.writer.add_scalar("Loss/VGG", errors["VGG"].mean().float(), step)
self.writer.add_scalars(
"Loss/GAN",
{
"G": errors["GAN"].mean().float(),
"D": (errors["D_Fake"].mean().float() + errors["D_real"].mean().float()) / 2,
},
step,
)
# errors: same format as |errors| of plotCurrentErrors
def print_current_errors(self, epoch, i, errors, t):
message = "(epoch: %d, iters: %d, time: %.3f) " % (epoch, i, t)
for k, v in errors.items():
v = v.mean().float()
message += "%s: %.3f " % (k, v)
print(message)
with open(self.log_name, "a") as log_file:
log_file.write("%s\n" % message)
def convert_visuals_to_numpy(self, visuals):
for key, t in visuals.items():
tile = self.opt.batchSize > 8
if "input_label" == key:
t = util.tensor2label(t, self.opt.label_nc + 2, tile=tile) ## B*H*W*C 0-255 numpy
else:
t = util.tensor2im(t, tile=tile)
visuals[key] = t
return visuals
# save image to the disk
def save_images(self, webpage, visuals, image_path):
visuals = self.convert_visuals_to_numpy(visuals)
image_dir = webpage.get_image_dir()
short_path = ntpath.basename(image_path[0])
name = os.path.splitext(short_path)[0]
webpage.add_header(name)
ims = []
txts = []
links = []
for label, image_numpy in visuals.items():
image_name = os.path.join(label, "%s.png" % (name))
save_path = os.path.join(image_dir, image_name)
util.save_image(image_numpy, save_path, create_dir=True)
ims.append(image_name)
txts.append(label)
links.append(image_name)
webpage.add_images(ims, txts, links, width=self.win_size)
class Visualizer():
def __init__(self, opt):
self.opt = opt
self.tf_log = opt.isTrain and opt.tf_log
self.tensorboard_log = opt.tensorboard_log
self.use_html = opt.isTrain and not opt.no_html
self.win_size = opt.display_winsize
self.name = opt.name
if self.tensorboard_log:
if self.opt.isTrain:
self.log_dir = os.path.join(opt.checkpoints_dir, opt.name,
'logs')
if not os.path.exists(self.log_dir):
os.makedirs(self.log_dir)
self.writer = SummaryWriter(log_dir=self.log_dir)
else:
print("hi :)")
self.log_dir = os.path.join(opt.checkpoints_dir, opt.name,
opt.results_dir)
if not os.path.exists(self.log_dir):
os.makedirs(self.log_dir)
# import tensorflow as tf
# self.tf = tf
# self.log_dir = os.path.join(opt.checkpoints_dir, opt.name, 'logs')
# self.writer = tf.summary.FileWriter(self.log_dir)
if opt.isTrain:
self.log_name = os.path.join(opt.checkpoints_dir, opt.name,
'loss_log.txt')
with open(self.log_name, "a") as log_file:
now = time.strftime("%c")
log_file.write(
'================ Training Loss (%s) ================\n' %
now)
# |visuals|: dictionary of images to display or save
def display_current_results(self, visuals, epoch, step):
all_tensor = []
if self.tensorboard_log:
for key, tensor in visuals.items():
if key == 'input_label':
tile = self.opt.batchSize > 1
t = util.tensor2label(tensor,
self.opt.label_nc + 2,
tile=tile) ## B*H*W*3 0-255 numpy
t = np.transpose(t, (0, 3, 1, 2))
all_tensor.append(torch.tensor(t).float() / 255)
else:
all_tensor.append((tensor.data.cpu() + 1) / 2)
output = torch.cat(all_tensor, 0)
img_grid = vutils.make_grid(output,
nrow=self.opt.batchSize,
padding=0,
normalize=False)
if self.opt.isTrain:
self.writer.add_image('Face_SPADE/training_samples', img_grid,
step)
else:
# self.writer.add_image('Face_SPADE/test_samples',img_grid,step)
vutils.save_image(output,
os.path.join(self.log_dir,
str(step) + '.png'),
nrow=self.opt.batchSize,
padding=0,
normalize=False)
## convert tensors to numpy arrays
# visuals = self.convert_visuals_to_numpy(visuals)
# if self.tf_log: # show images in tensorboard output
# img_summaries = []
# for label, image_numpy in visuals.items():
# # Write the image to a string
# try:
# s = StringIO()
# except:
# s = BytesIO()
# if len(image_numpy.shape) >= 4:
# image_numpy = image_numpy[0]
# scipy.misc.toimage(image_numpy).save(s, format="jpeg")
# # Create an Image object
# img_sum = self.tf.Summary.Image(encoded_image_string=s.getvalue(), height=image_numpy.shape[0], width=image_numpy.shape[1])
# # Create a Summary value
# img_summaries.append(self.tf.Summary.Value(tag=label, image=img_sum))
# # Create and write Summary
# summary = self.tf.Summary(value=img_summaries)
# self.writer.add_summary(summary, step)
# errors: dictionary of error labels and values
def plot_current_errors(self, errors, step):
if self.tf_log:
for tag, value in errors.items():
value = value.mean().float()
summary = self.tf.Summary(
value=[self.tf.Summary.Value(tag=tag, simple_value=value)])
self.writer.add_summary(summary, step)
if self.tensorboard_log:
# self.writer.add_scalar('G',errors['GAN'].item(),step)
self.writer.add_scalar('Loss/GAN_Feat',
errors['GAN_Feat'].mean().float(), step)
self.writer.add_scalar('Loss/VGG', errors['VGG'].mean().float(),
step)
# self.writer.add_scalar('D',(errors['D_Fake'].item()+errors['D_real'].item())/2,step)
self.writer.add_scalars(
'Loss/GAN', {
'G':
errors['GAN'].mean().float(),
'D': (errors['D_Fake'].mean().float() +
errors['D_real'].mean().float()) / 2
}, step)
# errors: same format as |errors| of plotCurrentErrors
def print_current_errors(self, epoch, i, errors, t):
message = '(epoch: %d, iters: %d, time: %.3f) ' % (epoch, i, t)
for k, v in errors.items():
#print(v)
#if v != 0:
v = v.mean().float()
message += '%s: %.3f ' % (k, v)
print(message)
with open(self.log_name, "a") as log_file:
log_file.write('%s\n' % message)
def convert_visuals_to_numpy(self, visuals):
for key, t in visuals.items():
tile = self.opt.batchSize > 8
if 'input_label' == key:
t = util.tensor2label(t, self.opt.label_nc + 2,
tile=tile) ## B*H*W*C 0-255 numpy
else:
t = util.tensor2im(t, tile=tile)
visuals[key] = t
return visuals
# save image to the disk
def save_images(self, webpage, visuals, image_path):
visuals = self.convert_visuals_to_numpy(visuals)
image_dir = webpage.get_image_dir()
short_path = ntpath.basename(image_path[0])
name = os.path.splitext(short_path)[0]
webpage.add_header(name)
ims = []
txts = []
links = []
for label, image_numpy in visuals.items():
image_name = os.path.join(label, '%s.png' % (name))
save_path = os.path.join(image_dir, image_name)
util.save_image(image_numpy, save_path, create_dir=True)
ims.append(image_name)
txts.append(label)
links.append(image_name)
webpage.add_images(ims, txts, links, width=self.win_size)
