File size: 11,164 Bytes
51cbdf4 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 | import os
import logging
import itertools
import numpy as np
import random
import einops
import torch
import torchvision
from PIL import Image
from pytorch_lightning.callbacks import Callback
from pytorch_lightning.loggers import TensorBoardLogger
from pytorch_lightning.utilities.distributed import rank_zero_only
import matplotlib.pyplot as plt
class ImageLogger(Callback):
def __init__(self, val_dataloader=None, batch_frequency=2000, max_images=4, clamp=True, increase_log_steps=True,
rescale=True, disabled=False, log_on_batch_idx=False, log_first_step=False,
log_images_kwargs=None, experiment=None, plot_frequency=300, num_samples=1):
super().__init__()
self.rescale = rescale
self.batch_freq = batch_frequency
self.max_images = max_images
if not increase_log_steps:
self.log_steps = [self.batch_freq]
self.clamp = clamp
self.disabled = disabled
self.log_on_batch_idx = log_on_batch_idx
self.log_images_kwargs = log_images_kwargs if log_images_kwargs else {}
self.log_first_step = log_first_step
self.experiment = experiment
self.log_path = os.path.join("log", "image_log_{}".format(self.experiment))
self.init_loss_logger()
self.training_losses = []
self.global_steps = []
self.plot_frequency = plot_frequency
self.val_dataloader = val_dataloader
self.num_samples = num_samples
def init_loss_logger(self):
os.makedirs(self.log_path, exist_ok=True)
logging.basicConfig(filename=os.path.join(self.log_path, "training_loss.log"), level=logging.INFO,
format='%(asctime)s - %(message)s', datefmt='%Y-%m-%d %H:%M:%S')
@rank_zero_only
def log_loss(self, loss, global_step):
logging.info(f"Step: {global_step}, Loss: {loss}")
self.training_losses.append(loss)
self.global_steps.append(global_step)
@rank_zero_only
def update_loss_plot(self):
plt.figure()
plt.plot(self.global_steps, self.training_losses, label="Training loss")
plt.xlabel("Global Step")
plt.ylabel("Loss")
plt.legend()
plt.savefig(os.path.join(self.log_path, "training_loss_plot.png"))
plt.close()
@rank_zero_only
def log_local(self, save_dir, split, images, global_step, current_epoch, batch_idx):
root = os.path.join(save_dir, "log", "image_log_{}".format(self.experiment), split)
for k in images:
grid = torchvision.utils.make_grid(images[k], nrow=4)
if self.rescale:
grid = (grid + 1.0) / 2.0 # -1,1 -> 0,1; c,h,w
grid = grid.transpose(0, 1).transpose(1, 2).squeeze(-1)
grid = grid.numpy()
grid = (grid * 255).astype(np.uint8)
filename = "{}_gs-{:06}_e-{:06}_b-{:06}.png".format(k, global_step, current_epoch, batch_idx)
path = os.path.join(root, filename)
os.makedirs(os.path.split(path)[0], exist_ok=True)
Image.fromarray(grid).save(path)
@rank_zero_only
def log_local_val(self, save_dir, split, images, global_step, current_epoch, batch_idx):
root = os.path.join(save_dir, "log", "image_log_{}".format(self.experiment), split, str(current_epoch))
for k in images:
for idx, image in enumerate(images[k]):
if self.rescale:
image = (image + 1.0) / 2.0 # -1,1 -> 0,1; c,h,w
image = image.transpose(0, 1).transpose(1, 2).squeeze(-1)
image = image.numpy()
image = (image * 255).astype(np.uint8)
filename = "{}_gs-{:06}_e-{:06}_b-{:06}_i-{:06}.png".format(k, global_step, current_epoch, batch_idx, idx)
path = os.path.join(root, filename)
os.makedirs(os.path.split(path)[0], exist_ok=True)
Image.fromarray(image).save(path)
def log_img(self, pl_module, batch, batch_idx, split="train"):
check_idx = batch_idx # if self.log_on_batch_idx else pl_module.global_step
if (self.check_frequency(check_idx) and # batch_idx % self.batch_freq == 0
hasattr(pl_module, "log_images") and
callable(pl_module.log_images) and
self.max_images > 0):
logger = type(pl_module.logger)
is_train = pl_module.training
if is_train:
pl_module.eval()
with torch.no_grad():
images = pl_module.log_images(batch, split=split, **self.log_images_kwargs)
for k in images:
N = min(images[k].shape[0], self.max_images)
images[k] = images[k][:N]
if isinstance(images[k], torch.Tensor):
images[k] = images[k].detach().cpu()
if self.clamp:
images[k] = torch.clamp(images[k], -1., 1.)
self.log_local(pl_module.logger.save_dir, split, images,
pl_module.global_step, pl_module.current_epoch, batch_idx)
if is_train:
pl_module.train()
def log_img_val(self, pl_module, batch, batch_idx, split="val"):
check_idx = batch_idx # if self.log_on_batch_idx else pl_module.global_step
if (hasattr(pl_module, "log_images") and callable(pl_module.log_images) and self.max_images > 0):
logger = type(pl_module.logger)
is_train = pl_module.training
if is_train:
pl_module.eval()
with torch.no_grad():
images = pl_module.log_images(batch, split=split, num_samples=self.num_samples, **self.log_images_kwargs)
for k in images:
N = min(images[k].shape[0], self.num_samples)
images[k] = images[k][:N]
if isinstance(images[k], torch.Tensor):
images[k] = images[k].detach().cpu()
if self.clamp:
images[k] = torch.clamp(images[k], -1., 1.)
self.log_local_val(pl_module.logger.save_dir, split, images,
pl_module.global_step, pl_module.current_epoch, batch_idx)
if is_train:
pl_module.train()
def check_frequency(self, check_idx):
return check_idx % self.batch_freq == 0
def on_train_batch_end(self, trainer, pl_module, outputs, batch, batch_idx, dataloader_idx):
if not self.disabled:
self.log_img(pl_module, batch, batch_idx, split="train")
'''
if "loss" in outputs:
train_loss = outputs["loss"]
if isinstance(train_loss, torch.Tensor):
train_loss = train_loss.detach().cpu().numpy()
self.log_loss(train_loss, pl_module.global_step)
if pl_module.global_step % self.plot_frequency == 0:
self.update_loss_plot()
'''
def on_epoch_end(self, trainer, pl_module):
if not self.disabled:
for batch_idx, batch in enumerate(self.val_dataloader):
input_image, prompt, hint = batch['jpg'], batch['txt'], batch['hint']
self.log_img_val(pl_module, batch, batch_idx, split="val")
'''
class ImageLogger(Callback):
def __init__(self, batch_frequency=2000, max_images=4, clamp=True, increase_log_steps=True,
rescale=True, disabled=False, log_on_batch_idx=False, log_first_step=False,
log_images_kwargs=None, experiment=None):
super().__init__()
self.rescale = rescale
self.batch_freq = batch_frequency
self.max_images = max_images
if not increase_log_steps:
self.log_steps = [self.batch_freq]
self.clamp = clamp
self.disabled = disabled
self.log_on_batch_idx = log_on_batch_idx
self.log_images_kwargs = log_images_kwargs if log_images_kwargs else {}
self.log_first_step = log_first_step
self.experiment = experiment
self.init_tensorboard_logger()
def init_tensorboard_logger(self):
self.logger = TensorBoardLogger("log/image_log_{}".format(self.experiment), name="my_experiment")
@rank_zero_only
def log_local(self, save_dir, split, images, global_step, current_epoch, batch_idx):
root = os.path.join(save_dir, "log", "image_log_{}".format(self.experiment), split)
for k in images:
grid = torchvision.utils.make_grid(images[k], nrow=4)
if self.rescale:
grid = (grid + 1.0) / 2.0 # -1,1 -> 0,1; c,h,w
grid = grid.transpose(0, 1).transpose(1, 2).squeeze(-1)
grid = grid.numpy()
grid = (grid * 255).astype(np.uint8)
filename = "{}_gs-{:06}_e-{:06}_b-{:06}.png".format(k, global_step, current_epoch, batch_idx)
path = os.path.join(root, filename)
os.makedirs(os.path.split(path)[0], exist_ok=True)
Image.fromarray(grid).save(path)
def log_img(self, pl_module, batch, batch_idx, split="train"):
check_idx = batch_idx # if self.log_on_batch_idx else pl_module.global_step
if (self.check_frequency(check_idx) and # batch_idx % self.batch_freq == 0
hasattr(pl_module, "log_images") and
callable(pl_module.log_images) and
self.max_images > 0):
logger = type(pl_module.logger)
is_train = pl_module.training
if is_train:
pl_module.eval()
with torch.no_grad():
images = pl_module.log_images(batch, split=split, **self.log_images_kwargs)
for k in images:
N = min(images[k].shape[0], self.max_images)
images[k] = images[k][:N]
if isinstance(images[k], torch.Tensor):
images[k] = images[k].detach().cpu()
if self.clamp:
images[k] = torch.clamp(images[k], -1., 1.)
self.log_local(pl_module.logger.save_dir, split, images,
pl_module.global_step, pl_module.current_epoch, batch_idx)
if is_train:
pl_module.train()
def log_hyperparams(self, *args, **kwargs):
self.logger.log_hyperparams(*args, **kwargs)
def log_graph(self, *args, **kwargs):
self.logger.log_graph(*args, **kwargs)
def save(self):
self.logger.save()
@property
def save_dir(self):
return self.logger.save_dir
def check_frequency(self, check_idx):
return check_idx % self.batch_freq == 0
def on_train_batch_end(self, trainer, pl_module, outputs, batch, batch_idx, dataloader_idx):
if not self.disabled:
self.log_img(pl_module, batch, batch_idx, split="train")
if "loss" in outputs:
train_loss = outputs["loss"]
if isinstance(train_loss, torch.Tensor):
train_loss = train_loss.detach().cpu().numpy()
self.logger.experiment.add_scalar("Loss/train", train_loss, pl_module.global_step)
''' |