Spaces:
Running
on
Zero
Running
on
Zero
| import os | |
| import numpy as np | |
| from PIL import Image | |
| import torch | |
| import torchvision | |
| from PIL import Image | |
| from pytorch_lightning.callbacks import Callback | |
| try: | |
| from pytorch_lightning.utilities.distributed import rank_zero_only | |
| except: | |
| from pytorch_lightning.utilities.rank_zero import rank_zero_only | |
| 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, num_local_conditions=7): | |
| 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.num_local_conditions = num_local_conditions | |
| def log_local(self, save_dir, split, images, global_step, current_epoch, batch_idx): | |
| root = os.path.join(save_dir, "image_log", split) | |
| for k in images: | |
| if k == 'local_control': | |
| _, chn, h, w = images[k].shape | |
| if h == w == 1: | |
| continue | |
| for local_idx in range(chn//3): | |
| grid = torchvision.utils.make_grid(images[k][:, 3*local_idx: 3*(local_idx+1), :, : ], 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(global_step, current_epoch, batch_idx, k, local_idx) | |
| path = os.path.join(root, filename) | |
| os.makedirs(os.path.split(path)[0], exist_ok=True) | |
| Image.fromarray(grid).save(path) | |
| elif k != 'global_control': | |
| 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(global_step, current_epoch, batch_idx, k) | |
| 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 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): | |
| if not self.disabled: | |
| self.log_img(pl_module, batch, batch_idx, split="train") | |