test problem

[liuyingbin123]

AttributeError: Can't pickle local object 'DeepWBNet.init..'
运行test.py提示无法序列化这个类。
另外想问下你的python版本那是多少

[onpix]

可以提供完整的报错信息吗？我的python版本是3.8。

[liuyingbin123]

命令：HYDRA_FULL_ERROR=1 python src/test.py checkpoint_path=trained_on_ours.ckpt backend=cuda

报错信息：
Global seed set to 233
src/test.py:15: UserWarning:
The version_base parameter is not specified.
Please specify a compatability version level, or None.
Will assume defaults for version 1.1
@hydra.main(config_path='config', config_name="config")
/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'config': Defaults list is missing _self_. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information
warnings.warn(msg, UserWarning)
/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
ret = run_job(
Check runtime config: use "/data/lyb/code/LCDPNet/src/config/runtime/lcdpnet.default.yaml" as template.
Running config: {'aug': {'crop': False, 'downsample': [512, 512], 'h-flip': True, 'v-flip': True}, 'train_ds': {'class': 'img_dataset', 'name': 'lcdp_data.train', 'input': ['/data/lyb/data/lowlight/LCDP/data/input/'], 'GT': ['/data/lyb/data/lowlight/LCDP/data/gt/']}, 'test_ds': {'class': 'img_dataset', 'name': 'lcdp_data.test', 'input': ['/data/lyb/data/lowlight/LCDP/data/test-input/'], 'GT': ['/data/lyb/data/lowlight/LCDP/data/test-gt/']}, 'valid_ds': {'class': 'img_dataset', 'name': 'lcdp_data.valid', 'input': ['/data/lyb/data/lowlight/LCDP/data/valid-input/'], 'GT': ['/data/lyb/data/lowlight/LCDP/data/valid-gt/']}, 'runtime': {'bilateral_upsample_net': {'hist_unet': {'n_bins': 8, 'hist_as_guide': False, 'channel_nums': [8, 16, 32, 64, 128], 'encoder_use_hist': False, 'guide_feature_from_hist': True, 'region_num': 2, 'use_gray_hist': False, 'conv_type': 'drconv', 'down_ratio': 2, 'hist_conv_trainable': False, 'drconv_position': [0, 1]}, 'modelname': 'bilateral_upsample_net', 'predict_illumination': False, 'loss': {'mse': 1.0, 'cos': 0.1, 'ltv': 0.1}, 'luma_bins': 8, 'channel_multiplier': 1, 'spatial_bin': 16, 'batch_norm': True, 'low_resolution': 256, 'coeffs_type': 'matrix', 'conv_type': 'conv', 'backbone': 'hist-unet', 'illu_map_power': False}, 'hist_unet': {'n_bins': 8, 'hist_as_guide': False, 'channel_nums': False, 'encoder_use_hist': False, 'guide_feature_from_hist': False, 'region_num': 8, 'use_gray_hist': False, 'conv_type': 'drconv', 'down_ratio': 2, 'hist_conv_trainable': False, 'drconv_position': [1, 1]}, 'modelname': 'lcdpnet', 'use_wavelet': False, 'use_attn_map': False, 'use_non_local': False, 'how_to_fuse': 'cnn-weights', 'backbone': 'bilateral_upsample_net', 'conv_type': 'conv', 'backbone_out_illu': True, 'illumap_channel': 3, 'share_weights': True, 'n_bins': 8, 'hist_as_guide': False, 'loss': {'ltv': 0, 'cos': 0, 'weighted_loss': 0, 'tvloss1': 0, 'tvloss2': 0, 'tvloss1_new': 0.01, 'tvloss2_new': 0.01, 'l1_loss': 1.0, 'ssim_loss': 0, 'psnr_loss': 0, 'illumap_loss': 0, 'hist_loss': 0, 'inter_hist_loss': 0, 'vgg_loss': 0, 'cos2': 0.5}}, 'project': 'default_proj', 'name': 'default_name', 'comment': False, 'debug': False, 'val_debug_step_nums': 2, 'gpu': -1, 'backend': 'cuda', 'runtime_precision': 16, 'amp_backend': 'native', 'amp_level': 'O1', 'dataloader_num_worker': 5, 'mode': 'train', 'logger': 'tb', 'num_epoch': 1000, 'valid_every': 10, 'savemodel_every': 4, 'log_every': 100, 'batchsize': 1, 'valid_batchsize': 1, 'lr': 0.0001, 'checkpoint_path': 'trained_on_ours.ckpt', 'checkpoint_monitor': 'loss', 'resume_training': True, 'monitor_mode': 'min', 'early_stop': False, 'valid_ratio': 0.1, 'flags': {}}
ERR: import thop failed, skip. error msg:
No module named 'thop'
Lightning automatically upgraded your loaded checkpoint from v1.5.4 to v1.9.0. To apply the upgrade to your files permanently, run python -m pytorch_lightning.utilities.upgrade_checkpoint --file trained_on_ours.ckpt
[ WARN ] Use Conv in HistGuidedDRDoubleConv[0] instead of DRconv.
[ WARN ] Use Conv in HistGuidedDRDoubleConv[0] instead of DRconv.
[ WARN ] Use Conv in HistGuidedDRDoubleConv[0] instead of DRconv.
[ WARN ] Use Conv in HistGuidedDRDoubleConv[0] instead of DRconv.
[[ WARN ]] Using HistUNet in BilateralUpsampleNet as backbone
Running initialization for BaseModel
DeepWBNet(
(illu_net): BilateralUpsampleNet(
(guide): GuideNet(
(conv1): ConvBlock(
(conv): Conv2d(3, 16, kernel_size=(1, 1), stride=(1, 1))
(activation): ReLU()
(bn): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
(conv2): ConvBlock(
(conv): Conv2d(16, 1, kernel_size=(1, 1), stride=(1, 1))
(activation): Sigmoid()
)
)
(slice): SliceNode()
(coeffs): LowResHistUNet(
(maxpool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(hist_conv): Conv2d(8, 8, kernel_size=(2, 2), stride=(2, 2), bias=False)
(inc): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(3, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(8, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
(down1): Down(
(maxpool_conv): Sequential(
(0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(1): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(8, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(16, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
)
)
(down2): Down(
(maxpool_conv): Sequential(
(0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(1): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(16, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
)
)
(down3): Down(
(maxpool_conv): Sequential(
(0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(1): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
)
)
(down4): Down(
(maxpool_conv): Sequential(
(0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(1): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
)
)
(up1): Up(
(up): Upsample(scale_factor=2.0, mode=bilinear)
(conv): HistGuidedDRDoubleConv(
(conv1): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(inter1): Sequential(
(0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
(conv2): DRConv2d(
(conv_kernel): Sequential(
(0): AdaptiveAvgPool2d(output_size=(3, 3))
(1): Conv2d(64, 4, kernel_size=(1, 1), stride=(1, 1))
(2): Sigmoid()
(3): Conv2d(4, 4096, kernel_size=(1, 1), stride=(1, 1), groups=2)
)
(conv_guide): Conv2d(24, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(corr): Correlation(xcorr_fast)
)
(inter2): Sequential(
(0): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
)
)
(up2): Up(
(up): Upsample(scale_factor=2.0, mode=bilinear)
(conv): HistGuidedDRDoubleConv(
(conv1): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(inter1): Sequential(
(0): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
(conv2): DRConv2d(
(conv_kernel): Sequential(
(0): AdaptiveAvgPool2d(output_size=(3, 3))
(1): Conv2d(32, 4, kernel_size=(1, 1), stride=(1, 1))
(2): Sigmoid()
(3): Conv2d(4, 1024, kernel_size=(1, 1), stride=(1, 1), groups=2)
)
(conv_guide): Conv2d(24, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(corr): Correlation(xcorr_fast)
)
(inter2): Sequential(
(0): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
)
)
(up3): Up(
(up): Upsample(scale_factor=2.0, mode=bilinear)
(conv): HistGuidedDRDoubleConv(
(conv1): Conv2d(32, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(inter1): Sequential(
(0): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
(conv2): DRConv2d(
(conv_kernel): Sequential(
(0): AdaptiveAvgPool2d(output_size=(3, 3))
(1): Conv2d(16, 4, kernel_size=(1, 1), stride=(1, 1))
(2): Sigmoid()
(3): Conv2d(4, 256, kernel_size=(1, 1), stride=(1, 1), groups=2)
)
(conv_guide): Conv2d(24, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(corr): Correlation(xcorr_fast)
)
(inter2): Sequential(
(0): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
)
)
(up4): Up(
(up): Upsample(scale_factor=2.0, mode=bilinear)
(conv): HistGuidedDRDoubleConv(
(conv1): Conv2d(16, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(inter1): Sequential(
(0): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
(conv2): DRConv2d(
(conv_kernel): Sequential(
(0): AdaptiveAvgPool2d(output_size=(3, 3))
(1): Conv2d(8, 4, kernel_size=(1, 1), stride=(1, 1))
(2): Sigmoid()
(3): Conv2d(4, 128, kernel_size=(1, 1), stride=(1, 1), groups=2)
)
(conv_guide): Conv2d(24, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(corr): Correlation(xcorr_fast)
)
(inter2): Sequential(
(0): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
)
)
(outc): OutConv(
(conv): Conv2d(8, 96, kernel_size=(1, 1), stride=(1, 1))
)
)
(apply_coeffs): ApplyCoeffs()
)
(out_net): Sequential(
(0): Conv2d(9, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): ReLU(inplace=True)
(2): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(3): ReLU(inplace=True)
(4): NONLocalBlock2D(
(g): Sequential(
(0): Conv2d(32, 16, kernel_size=(1, 1), stride=(1, 1))
(1): UpsamplingBilinear2d(size=[16, 16], mode=bilinear)
)
(W): Conv2d(16, 32, kernel_size=(1, 1), stride=(1, 1))
(theta): Conv2d(32, 16, kernel_size=(1, 1), stride=(1, 1))
(phi): Sequential(
(0): Conv2d(32, 16, kernel_size=(1, 1), stride=(1, 1))
(1): UpsamplingBilinear2d(size=[16, 16], mode=bilinear)
)
)
(5): Conv2d(32, 32, kernel_size=(1, 1), stride=(1, 1))
(6): ReLU(inplace=True)
(7): Conv2d(32, 3, kernel_size=(1, 1), stride=(1, 1))
(8): NONLocalBlock2D(
(g): Sequential(
(0): Conv2d(3, 1, kernel_size=(1, 1), stride=(1, 1))
(1): UpsamplingBilinear2d(size=[16, 16], mode=bilinear)
)
(W): Conv2d(1, 3, kernel_size=(1, 1), stride=(1, 1))
(theta): Conv2d(3, 1, kernel_size=(1, 1), stride=(1, 1))
(phi): Sequential(
(0): Conv2d(3, 1, kernel_size=(1, 1), stride=(1, 1))
(1): UpsamplingBilinear2d(size=[16, 16], mode=bilinear)
)
)
)
)
[ WARN ] Result directory "lcdpnet__trained_on_ours.ckpt@lcdp_data.test" exists. Press ENTER to overwrite or input suffix to create a new one:

New name: lcdpnet__trained_on_ours.ckpt@lcdp_data.test.
[ WARN ] Overwrite result_dir: lcdpnet__trained_on_ours.ckpt@lcdp_data.test
TEST - Result save path:
test_result/lcdpnet__trained_on_ours.ckpt@lcdp_data.test
Loading model from: trained_on_ours.ckpt
Dataset augmentation:
[ToPILImage(), Downsample([512, 512]), RandomHorizontalFlip(p=0.5), RandomVerticalFlip(p=0.5), ToTensor()]
/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/pytorch_lightning/trainer/connectors/accelerator_connector.py:467: LightningDeprecationWarning: Setting Trainer(gpus=-1) is deprecated in v1.7 and will be removed in v2.0. Please use Trainer(accelerator='gpu', devices=-1) instead.
rank_zero_deprecation(
Using 16bit None Automatic Mixed Precision (AMP)
GPU available: True (cuda), used: True
TPU available: False, using: 0 TPU cores
IPU available: False, using: 0 IPUs
HPU available: False, using: 0 HPUs
Error executing job with overrides: ['checkpoint_path=trained_on_ours.ckpt', 'backend=cuda']
Traceback (most recent call last):
File "src/test.py", line 46, in
main()
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/main.py", line 90, in decorated_main
_run_hydra(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra
_run_app(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 457, in _run_app
run_and_report(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 222, in run_and_report
raise ex
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 219, in run_and_report
return func()
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 458, in
lambda: hydra.run(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/hydra.py", line 132, in run
_ = ret.return_value
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/core/utils.py", line 260, in return_value
raise self._return_value
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/core/utils.py", line 186, in run_job
ret.return_value = task_function(task_cfg)
File "src/test.py", line 39, in main
trainer.test(model, datamodule)
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 785, in test
return call._call_and_handle_interrupt(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/pytorch_lightning/trainer/call.py", line 36, in _call_and_handle_interrupt
return trainer.strategy.launcher.launch(trainer_fn, *args, trainer=trainer, **kwargs)
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/pytorch_lightning/strategies/launchers/multiprocessing.py", line 113, in launch
mp.start_processes(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 189, in start_processes
process.start()
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/process.py", line 121, in start
self._popen = self._Popen(self)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/context.py", line 284, in _Popen
return Popen(process_obj)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/popen_spawn_posix.py", line 32, in init
super().init(process_obj)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/popen_fork.py", line 19, in init
self._launch(process_obj)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/popen_spawn_posix.py", line 47, in _launch
reduction.dump(process_obj, fp)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/reduction.py", line 60, in dump
ForkingPickler(file, protocol).dump(obj)
TypeError: cannot pickle 'cv2.TonemapReinhard' object

我把模型里面的几个lambda替换成函数后就报和之前一个issue想通的错误了

[liuyingbin123]

或者老哥看下能不能提供下python、torch、还有requirements.txt里包的版本号

[onpix]

我刚刚测试了一下，我的环境下不做改动直接运行是没问题的。也许是你做了改动的缘故？目前我使用的是python3.8.8，torch相关环境是：

pytorch-lightning                  1.7.6
torch                              1.12.1
torchmetrics                       0.9.2
torchvision                        0.13.1

本代码发布时我使用的版本更老一些，不过二者都能正常运行。

[onpix]

若是由于修改代码导致的错误，我倾向认为是由于model代码修改后，model class和ckpt不匹配导致的。

[liuyingbin123]

不是的，修改前也不能直接运行，无法序列化带lambda的类

Global seed set to 233
src/test.py:15: UserWarning:
The version_base parameter is not specified.
Please specify a compatability version level, or None.
Will assume defaults for version 1.1
@hydra.main(config_path='config', config_name="config")
/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/defaults_list.py:251: UserWarning: In 'config': Defaults list is missing _self_. See https://hydra.cc/docs/1.2/upgrades/1.0_to_1.1/default_composition_order for more information
warnings.warn(msg, UserWarning)
/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
See https://hydra.cc/docs/1.2/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
ret = run_job(
Check runtime config: use "/data/lyb/code/LCDPNet/src/config/runtime/lcdpnet.default.yaml" as template.
Running config: {'aug': {'crop': False, 'downsample': [512, 512], 'h-flip': True, 'v-flip': True}, 'train_ds': {'class': 'img_dataset', 'name': 'lcdp_data.train', 'input': ['/data/lyb/data/lowlight/LCDP/data/input/'], 'GT': ['/data/lyb/data/lowlight/LCDP/data/gt/']}, 'test_ds': {'class': 'img_dataset', 'name': 'lcdp_data.test', 'input': ['/data/lyb/data/lowlight/LCDP/data/test-input/'], 'GT': ['/data/lyb/data/lowlight/LCDP/data/test-gt/']}, 'valid_ds': {'class': 'img_dataset', 'name': 'lcdp_data.valid', 'input': ['/data/lyb/data/lowlight/LCDP/data/valid-input/'], 'GT': ['/data/lyb/data/lowlight/LCDP/data/valid-gt/']}, 'runtime': {'bilateral_upsample_net': {'hist_unet': {'n_bins': 8, 'hist_as_guide': False, 'channel_nums': [8, 16, 32, 64, 128], 'encoder_use_hist': False, 'guide_feature_from_hist': True, 'region_num': 2, 'use_gray_hist': False, 'conv_type': 'drconv', 'down_ratio': 2, 'hist_conv_trainable': False, 'drconv_position': [0, 1]}, 'modelname': 'bilateral_upsample_net', 'predict_illumination': False, 'loss': {'mse': 1.0, 'cos': 0.1, 'ltv': 0.1}, 'luma_bins': 8, 'channel_multiplier': 1, 'spatial_bin': 16, 'batch_norm': True, 'low_resolution': 256, 'coeffs_type': 'matrix', 'conv_type': 'conv', 'backbone': 'hist-unet', 'illu_map_power': False}, 'hist_unet': {'n_bins': 8, 'hist_as_guide': False, 'channel_nums': False, 'encoder_use_hist': False, 'guide_feature_from_hist': False, 'region_num': 8, 'use_gray_hist': False, 'conv_type': 'drconv', 'down_ratio': 2, 'hist_conv_trainable': False, 'drconv_position': [1, 1]}, 'modelname': 'lcdpnet', 'use_wavelet': False, 'use_attn_map': False, 'use_non_local': False, 'how_to_fuse': 'cnn-weights', 'backbone': 'bilateral_upsample_net', 'conv_type': 'conv', 'backbone_out_illu': True, 'illumap_channel': 3, 'share_weights': True, 'n_bins': 8, 'hist_as_guide': False, 'loss': {'ltv': 0, 'cos': 0, 'weighted_loss': 0, 'tvloss1': 0, 'tvloss2': 0, 'tvloss1_new': 0.01, 'tvloss2_new': 0.01, 'l1_loss': 1.0, 'ssim_loss': 0, 'psnr_loss': 0, 'illumap_loss': 0, 'hist_loss': 0, 'inter_hist_loss': 0, 'vgg_loss': 0, 'cos2': 0.5}}, 'project': 'default_proj', 'name': 'default_name', 'comment': False, 'debug': False, 'val_debug_step_nums': 2, 'gpu': -1, 'backend': 'cuda', 'runtime_precision': 16, 'amp_backend': 'native', 'amp_level': 'O1', 'dataloader_num_worker': 5, 'mode': 'train', 'logger': 'tb', 'num_epoch': 1000, 'valid_every': 10, 'savemodel_every': 4, 'log_every': 100, 'batchsize': 1, 'valid_batchsize': 1, 'lr': 0.0001, 'checkpoint_path': 'trained_on_ours.ckpt', 'checkpoint_monitor': 'loss', 'resume_training': True, 'monitor_mode': 'min', 'early_stop': False, 'valid_ratio': 0.1, 'flags': {}}
ERR: import thop failed, skip. error msg:
No module named 'thop'
Lightning automatically upgraded your loaded checkpoint from v1.5.4 to v1.9.0. To apply the upgrade to your files permanently, run python -m pytorch_lightning.utilities.upgrade_checkpoint --file trained_on_ours.ckpt
[ WARN ] Use Conv in HistGuidedDRDoubleConv[0] instead of DRconv.
[ WARN ] Use Conv in HistGuidedDRDoubleConv[0] instead of DRconv.
[ WARN ] Use Conv in HistGuidedDRDoubleConv[0] instead of DRconv.
[ WARN ] Use Conv in HistGuidedDRDoubleConv[0] instead of DRconv.
[[ WARN ]] Using HistUNet in BilateralUpsampleNet as backbone
Running initialization for BaseModel
DeepWBNet(
(illu_net): BilateralUpsampleNet(
(guide): GuideNet(
(conv1): ConvBlock(
(conv): Conv2d(3, 16, kernel_size=(1, 1), stride=(1, 1))
(activation): ReLU()
(bn): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
)
(conv2): ConvBlock(
(conv): Conv2d(16, 1, kernel_size=(1, 1), stride=(1, 1))
(activation): Sigmoid()
)
)
(slice): SliceNode()
(coeffs): LowResHistUNet(
(maxpool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(hist_conv): Conv2d(8, 8, kernel_size=(2, 2), stride=(2, 2), bias=False)
(inc): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(3, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(8, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
(down1): Down(
(maxpool_conv): Sequential(
(0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(1): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(8, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(16, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
)
)
(down2): Down(
(maxpool_conv): Sequential(
(0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(1): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(16, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
)
)
(down3): Down(
(maxpool_conv): Sequential(
(0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(1): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
)
)
(down4): Down(
(maxpool_conv): Sequential(
(0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(1): DoubleConv(
(double_conv): Sequential(
(0): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): ReLU(inplace=True)
(3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): ReLU(inplace=True)
)
)
)
)
(up1): Up(
(up): Upsample(scale_factor=2.0, mode=bilinear)
(conv): HistGuidedDRDoubleConv(
(conv1): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(inter1): Sequential(
(0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
(conv2): DRConv2d(
(conv_kernel): Sequential(
(0): AdaptiveAvgPool2d(output_size=(3, 3))
(1): Conv2d(64, 4, kernel_size=(1, 1), stride=(1, 1))
(2): Sigmoid()
(3): Conv2d(4, 4096, kernel_size=(1, 1), stride=(1, 1), groups=2)
)
(conv_guide): Conv2d(24, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(corr): Correlation(xcorr_fast)
)
(inter2): Sequential(
(0): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
)
)
(up2): Up(
(up): Upsample(scale_factor=2.0, mode=bilinear)
(conv): HistGuidedDRDoubleConv(
(conv1): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(inter1): Sequential(
(0): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
(conv2): DRConv2d(
(conv_kernel): Sequential(
(0): AdaptiveAvgPool2d(output_size=(3, 3))
(1): Conv2d(32, 4, kernel_size=(1, 1), stride=(1, 1))
(2): Sigmoid()
(3): Conv2d(4, 1024, kernel_size=(1, 1), stride=(1, 1), groups=2)
)
(conv_guide): Conv2d(24, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(corr): Correlation(xcorr_fast)
)
(inter2): Sequential(
(0): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
)
)
(up3): Up(
(up): Upsample(scale_factor=2.0, mode=bilinear)
(conv): HistGuidedDRDoubleConv(
(conv1): Conv2d(32, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(inter1): Sequential(
(0): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
(conv2): DRConv2d(
(conv_kernel): Sequential(
(0): AdaptiveAvgPool2d(output_size=(3, 3))
(1): Conv2d(16, 4, kernel_size=(1, 1), stride=(1, 1))
(2): Sigmoid()
(3): Conv2d(4, 256, kernel_size=(1, 1), stride=(1, 1), groups=2)
)
(conv_guide): Conv2d(24, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(corr): Correlation(xcorr_fast)
)
(inter2): Sequential(
(0): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
)
)
(up4): Up(
(up): Upsample(scale_factor=2.0, mode=bilinear)
(conv): HistGuidedDRDoubleConv(
(conv1): Conv2d(16, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(inter1): Sequential(
(0): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
(conv2): DRConv2d(
(conv_kernel): Sequential(
(0): AdaptiveAvgPool2d(output_size=(3, 3))
(1): Conv2d(8, 4, kernel_size=(1, 1), stride=(1, 1))
(2): Sigmoid()
(3): Conv2d(4, 128, kernel_size=(1, 1), stride=(1, 1), groups=2)
)
(conv_guide): Conv2d(24, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(corr): Correlation(xcorr_fast)
)
(inter2): Sequential(
(0): BatchNorm2d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(1): ReLU(inplace=True)
)
)
)
(outc): OutConv(
(conv): Conv2d(8, 96, kernel_size=(1, 1), stride=(1, 1))
)
)
(apply_coeffs): ApplyCoeffs()
)
(out_net): Sequential(
(0): Conv2d(9, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): ReLU(inplace=True)
(2): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(3): ReLU(inplace=True)
(4): NONLocalBlock2D(
(g): Sequential(
(0): Conv2d(32, 16, kernel_size=(1, 1), stride=(1, 1))
(1): UpsamplingBilinear2d(size=[16, 16], mode=bilinear)
)
(W): Conv2d(16, 32, kernel_size=(1, 1), stride=(1, 1))
(theta): Conv2d(32, 16, kernel_size=(1, 1), stride=(1, 1))
(phi): Sequential(
(0): Conv2d(32, 16, kernel_size=(1, 1), stride=(1, 1))
(1): UpsamplingBilinear2d(size=[16, 16], mode=bilinear)
)
)
(5): Conv2d(32, 32, kernel_size=(1, 1), stride=(1, 1))
(6): ReLU(inplace=True)
(7): Conv2d(32, 3, kernel_size=(1, 1), stride=(1, 1))
(8): NONLocalBlock2D(
(g): Sequential(
(0): Conv2d(3, 1, kernel_size=(1, 1), stride=(1, 1))
(1): UpsamplingBilinear2d(size=[16, 16], mode=bilinear)
)
(W): Conv2d(1, 3, kernel_size=(1, 1), stride=(1, 1))
(theta): Conv2d(3, 1, kernel_size=(1, 1), stride=(1, 1))
(phi): Sequential(
(0): Conv2d(3, 1, kernel_size=(1, 1), stride=(1, 1))
(1): UpsamplingBilinear2d(size=[16, 16], mode=bilinear)
)
)
)
)
[ WARN ] Result directory "lcdpnet__trained_on_ours.ckpt@lcdp_data.test" exists. Press ENTER to overwrite or input suffix to create a new one:

New name: lcdpnet__trained_on_ours.ckpt@lcdp_data.test.
[ WARN ] Overwrite result_dir: lcdpnet__trained_on_ours.ckpt@lcdp_data.test
TEST - Result save path:
test_result/lcdpnet__trained_on_ours.ckpt@lcdp_data.test
Loading model from: trained_on_ours.ckpt
Dataset augmentation:
[ToPILImage(), Downsample([512, 512]), RandomHorizontalFlip(p=0.5), RandomVerticalFlip(p=0.5), ToTensor()]
/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/pytorch_lightning/trainer/connectors/accelerator_connector.py:467: LightningDeprecationWarning: Setting Trainer(gpus=-1) is deprecated in v1.7 and will be removed in v2.0. Please use Trainer(accelerator='gpu', devices=-1) instead.
rank_zero_deprecation(
Using 16bit None Automatic Mixed Precision (AMP)
GPU available: True (cuda), used: True
TPU available: False, using: 0 TPU cores
IPU available: False, using: 0 IPUs
HPU available: False, using: 0 HPUs
Error executing job with overrides: ['checkpoint_path=trained_on_ours.ckpt', 'backend=cuda']
Traceback (most recent call last):
File "src/test.py", line 46, in
main()
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/main.py", line 90, in decorated_main
_run_hydra(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra
_run_app(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 457, in _run_app
run_and_report(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 222, in run_and_report
raise ex
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 219, in run_and_report
return func()
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/utils.py", line 458, in
lambda: hydra.run(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/_internal/hydra.py", line 132, in run
_ = ret.return_value
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/core/utils.py", line 260, in return_value
raise self._return_value
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/hydra/core/utils.py", line 186, in run_job
ret.return_value = task_function(task_cfg)
File "src/test.py", line 39, in main
trainer.test(model, datamodule)
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 785, in test
return call._call_and_handle_interrupt(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/pytorch_lightning/trainer/call.py", line 36, in _call_and_handle_interrupt
return trainer.strategy.launcher.launch(trainer_fn, *args, trainer=trainer, **kwargs)
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/pytorch_lightning/strategies/launchers/multiprocessing.py", line 113, in launch
mp.start_processes(
File "/opt/anaconda3/envs/rvm/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 189, in start_processes
process.start()
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/process.py", line 121, in start
self._popen = self._Popen(self)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/context.py", line 284, in _Popen
return Popen(process_obj)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/popen_spawn_posix.py", line 32, in init
super().init(process_obj)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/popen_fork.py", line 19, in init
self._launch(process_obj)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/popen_spawn_posix.py", line 47, in _launch
reduction.dump(process_obj, fp)
File "/opt/anaconda3/envs/rvm/lib/python3.8/multiprocessing/reduction.py", line 60, in dump
ForkingPickler(file, protocol).dump(obj)
AttributeError: Can't pickle local object 'DeepWBNet.init..'

[liuyingbin123]

建议老哥再提供一下requirements.txt中的版本号，我确保和你那边的环境是一致的

[onpix]

requirements.txt已更新。顺便上传了env.yaml，可以直接使用conda安装。

[liuyingbin123]

感谢