[Bug] KeyError: 'gt_label'

yanshuangying888 commented 7 months ago

分支

main 分支 (mmpretrain 版本)

描述该错误

config.py内容如下：（The contents of config.py are as follows:）

auto_scale_lr = dict(base_batch_size=256)
data_preprocessor = dict(
    mean=[
        123.675,
        116.28,
        103.53,
    ],
    num_classes=7,
    std=[
        58.395,
        57.12,
        57.375,
    ],
    to_rgb=True)
dataset_type = 'ImageNet'
default_hooks = dict(
    checkpoint=dict(interval=50, type='CheckpointHook'),
    logger=dict(interval=100, type='LoggerHook'),
    param_scheduler=dict(type='ParamSchedulerHook'),
    sampler_seed=dict(type='DistSamplerSeedHook'),
    timer=dict(type='IterTimerHook'),
    visualization=dict(enable=False, type='VisualizationHook'))
default_scope = 'mmpretrain'
env_cfg = dict(
    cudnn_benchmark=False,
    dist_cfg=dict(backend='nccl'),
    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
launcher = 'none'
load_from = None
log_level = 'INFO'
model = dict(
    backbone=dict(
        depth=50,
        num_stages=4,
        out_indices=(3, ),
        style='pytorch',
        type='ResNet'),
    head=dict(
        in_channels=2048,
        loss=dict(loss_weight=1.0, type='CrossEntropyLoss'),
        num_classes=7,
        topk=(
            1,
            5,
        ),
        type='LinearClsHead'),
    neck=dict(type='GlobalAveragePooling'),
    type='ImageClassifier')
optim_wrapper = dict(
    optimizer=dict(lr=0.1, momentum=0.9, type='SGD', weight_decay=0.0001))
param_scheduler = dict(
    by_epoch=True,
    gamma=0.1,
    milestones=[
        100,
        200,
        300,
        400,
        500,
    ],
    type='MultiStepLR')
randomness = dict(deterministic=False, seed=None)
resume = False
test_cfg = dict()
test_dataloader = dict(
    batch_size=64,
    collate_fn=dict(type='default_collate'),
    dataset=dict(
        data_root='D:\\data',
        pipeline=[
            dict(type='LoadImageFromFile'),
            dict(edge='short', scale=256, type='ResizeEdge'),
            dict(crop_size=224, type='CenterCrop'),
            dict(type='PackInputs'),
        ],
        split='test',
        type='ImageNet'),
    num_workers=8,
    persistent_workers=True,
    pin_memory=True,
    sampler=dict(shuffle=False, type='DefaultSampler'))
test_evaluator = [
    dict(topk=(
        1,
        5,
    ), type='Accuracy'),
    dict(
        items=[
            'precision',
            'recall',
            'f1-score',
        ], type='SingleLabelMetric'),
]
test_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(edge='short', scale=256, type='ResizeEdge'),
    dict(crop_size=224, type='CenterCrop'),
    dict(type='PackInputs'),
]
train_cfg = dict(by_epoch=True, max_epochs=500, val_interval=1)
train_dataloader = dict(
    batch_size=64,
    collate_fn=dict(type='default_collate'),
    dataset=dict(
        data_root='D:\\data',
        pipeline=[
            dict(type='LoadImageFromFile'),
            dict(scale=224, type='RandomResizedCrop'),
            dict(direction='horizontal', prob=0.5, type='RandomFlip'),
            dict(type='PackInputs'),
        ],
        split='train',
        type='ImageNet'),
    num_workers=8,
    persistent_workers=True,
    pin_memory=True,
    sampler=dict(shuffle=True, type='DefaultSampler'))
train_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(scale=224, type='RandomResizedCrop'),
    dict(direction='horizontal', prob=0.5, type='RandomFlip'),
    dict(type='PackInputs'),
]
val_cfg = dict()
val_dataloader = dict(
    batch_size=64,
    collate_fn=dict(type='default_collate'),
    dataset=dict(
        data_root='D:\\data',
        pipeline=[
            dict(type='LoadImageFromFile'),
            dict(edge='short', scale=256, type='ResizeEdge'),
            dict(crop_size=224, type='CenterCrop'),
            dict(type='PackInputs'),
        ],
        split='val',
        type='ImageNet'),
    num_workers=8,
    persistent_workers=True,
    pin_memory=True,
    sampler=dict(shuffle=False, type='DefaultSampler'))
val_evaluator = [
    dict(topk=(
        1,
        5,
    ), type='Accuracy'),
    dict(
        items=[
            'precision',
            'recall',
            'f1-score',
        ], type='SingleLabelMetric'),
]
vis_backends = [
    dict(type='LocalVisBackend'),
]
visualizer = dict(
    type='UniversalVisualizer', vis_backends=[
        dict(type='LocalVisBackend'),
    ])
work_dir = 'D:\\resnet50_8xb32_in1k_epoch500'

执行导出.pkl文件命令报错：（An error occurred when exporting.pkl files:）

python tools/test.py config.py resnet50_8xb32_in1k_epoch500.pth --out results.pkl

报错如下：（The error is as follows:）
Traceback (most recent call last):
  File "D:\Pythonfiles\Undergraduate_Thesis\mmpretrain-main\mmpretrain-main\tools\test.py", line 193, in <module>
    main()
  File "D:\Pythonfiles\Undergraduate_Thesis\mmpretrain-main\mmpretrain-main\tools\test.py", line 186, in main
    metrics = runner.test()
              ^^^^^^^^^^^^^
  File "C:\ProgramData\anaconda3\Lib\site-packages\mmengine\runner\runner.py", line 1823, in test
    metrics = self.test_loop.run()  # type: ignore
              ^^^^^^^^^^^^^^^^^^^^
  File "C:\ProgramData\anaconda3\Lib\site-packages\mmengine\runner\loops.py", line 443, in run
    self.run_iter(idx, data_batch)
  File "C:\ProgramData\anaconda3\Lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "C:\ProgramData\anaconda3\Lib\site-packages\mmengine\runner\loops.py", line 463, in run_iter
    self.evaluator.process(data_samples=outputs, data_batch=data_batch)
  File "C:\ProgramData\anaconda3\Lib\site-packages\mmengine\evaluator\evaluator.py", line 60, in process
    metric.process(data_batch, _data_samples)
  File "d:\pythonfiles\undergraduate_thesis\mmpretrain-main\mmpretrain-main\mmpretrain\evaluation\metrics\single_label.py", line 157, in process
    result['gt_label'] = data_sample['gt_label'].cpu()
                         ~~~~~~~~~~~^^^^^^^^^^^^
KeyError: 'gt_label'

请问这么解决这个问题呢？感激不尽您的建议。（How do you solve this problem? Your advice is greatly appreciated.）

环境信息

{'sys.platform': 'win32', 'Python': '3.11.5 | packaged by Anaconda, Inc. | (main, Sep 11 2023, ' '13:26:23) [MSC v.1916 64 bit (AMD64)]', 'CUDA available': True, 'MUSA available': False, 'numpy_random_seed': 2147483648, 'GPU 0': 'NVIDIA GeForce RTX 4090', 'CUDA_HOME': 'C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.1', 'NVCC': 'Cuda compilation tools, release 12.1, V12.1.105', 'MSVC': '用于 x64 的 Microsoft (R) C/C++ 优化编译器 19.39.33521 版', 'GCC': 'n/a', 'PyTorch': '2.1.2+cu121', 'TorchVision': '0.16.2+cu121', 'OpenCV': '4.9.0', 'MMEngine': '0.10.3', 'MMCV': '2.1.0', 'MMPreTrain': '1.2.0+unknown'}

其他信息

No response

JMcarrot commented 7 months ago

I also met this error,how do you solve it?

yanshuangying888 commented 7 months ago

I also met this error,how do you solve it? @JMcarrot 你好，朋友，很高兴能够回复你，我现在将我修正后的内容发给你，希望能够帮助到您。（Hello, my friend, I am very glad to reply to you. Now I will send you the corrected content, and I hope it can help you.）

auto_scale_lr = dict(base_batch_size=256)
data_preprocessor = dict(
mean=[
123.675,
116.28,
103.53,
],
num_classes=7,
std=[
58.395,
57.12,
57.375,
],
to_rgb=True)
dataset_type = 'ImageNet'
default_hooks = dict(
checkpoint=dict(interval=1, type='CheckpointHook'),
logger=dict(interval=100, type='LoggerHook'),
param_scheduler=dict(type='ParamSchedulerHook'),
sampler_seed=dict(type='DistSamplerSeedHook'),
timer=dict(type='IterTimerHook'),
visualization=dict(enable=False, type='VisualizationHook'))
default_scope = 'mmpretrain'
env_cfg = dict(
cudnn_benchmark=False,
dist_cfg=dict(backend='nccl'),
mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
launcher = 'none'
load_from = 'D:\\epoch_50.pth'
log_level = 'INFO'
model = dict(
backbone=dict(
depth=50,
num_stages=4,
out_indices=(3, ),
style='pytorch',
type='ResNet'),
head=dict(
in_channels=2048,
loss=dict(loss_weight=1.0, type='CrossEntropyLoss'),
num_classes=7,
topk=(
1,
5,
),
type='LinearClsHead'),
neck=dict(type='GlobalAveragePooling'),
type='ImageClassifier')
optim_wrapper = dict(
optimizer=dict(lr=0.1, momentum=0.9, type='SGD', weight_decay=0.0001))
param_scheduler = dict(
by_epoch=True,
gamma=0.1,
milestones=[
100,
200,
300,
400,
500,
],
type='MultiStepLR')
randomness = dict(deterministic=False, seed=None)
resume = False
test_cfg = dict()
test_dataloader = dict(
batch_size=64,
collate_fn=dict(type='default_collate'),
dataset=dict(
data_prefix='test',
data_root='D:\\data',
pipeline=[
dict(type='LoadImageFromFile'),
dict(edge='short', scale=256, type='ResizeEdge'),
dict(crop_size=224, type='CenterCrop'),
dict(type='PackInputs'),
],
type='CustomDataset'),
num_workers=8,
persistent_workers=True,
pin_memory=True,
sampler=dict(shuffle=False, type='DefaultSampler'))
test_evaluator = [
dict(topk=(
1,
5,
), type='Accuracy'),
dict(
items=[
'precision',
'recall',
'f1-score',
], type='SingleLabelMetric'),
]
test_pipeline = [
dict(type='LoadImageFromFile'),
dict(edge='short', scale=256, type='ResizeEdge'),
dict(crop_size=224, type='CenterCrop'),
dict(type='PackInputs'),
]
train_cfg = dict(by_epoch=True, max_epochs=500, val_interval=1)
train_dataloader = dict(
batch_size=64,
collate_fn=dict(type='default_collate'),
dataset=dict(
data_prefix='train',
data_root='D:\\data',
pipeline=[
dict(type='LoadImageFromFile'),
dict(scale=224, type='RandomResizedCrop'),
dict(direction='horizontal', prob=0.5, type='RandomFlip'),
dict(type='PackInputs'),
],
type='CustomDataset'),
num_workers=8,
persistent_workers=True,
pin_memory=True,
sampler=dict(shuffle=True, type='DefaultSampler'))
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(scale=224, type='RandomResizedCrop'),
dict(direction='horizontal', prob=0.5, type='RandomFlip'),
dict(type='PackInputs'),
]
val_cfg = dict()
val_dataloader = dict(
batch_size=64,
collate_fn=dict(type='default_collate'),
dataset=dict(
data_prefix='val',
data_root='D:\\data',
pipeline=[
dict(type='LoadImageFromFile'),
dict(edge='short', scale=256, type='ResizeEdge'),
dict(crop_size=224, type='CenterCrop'),
dict(type='PackInputs'),
],
type='CustomDataset'),
num_workers=8,
persistent_workers=True,
pin_memory=True,
sampler=dict(shuffle=False, type='DefaultSampler'))
val_evaluator = [
dict(topk=(
1,
5,
), type='Accuracy'),
dict(
items=[
'precision',
'recall',
'f1-score',
], type='SingleLabelMetric'),
]
vis_backends = [
dict(type='LocalVisBackend'),
]
visualizer = dict(
type='UniversalVisualizer', vis_backends=[
dict(type='LocalVisBackend'),
])
work_dir = 'D:\\resnet50_8xb32_in1k_epoch500'

你可以重点关注data_prefix=XXX这一部分，将原有的split=XXX进行替换。（You can focus on the data_prefix=XXX section and replace the original split=XXX.）希望这能帮助到您。（Hope this helps you.）

Jayden-ch commented 6 months ago

@yanshuangying888 你好，我也遇到了同样的问题，按照您的方法修改了data_prefix后，依旧没有得到解决

yanshuangying888 commented 6 months ago

@yanshuangying888 你好，我也遇到了同样的问题，按照您的方法修改了data_prefix后，依旧没有得到解决

@Jayden-ch 您好，朋友，很高兴能够回复您，我在修正此问题时，参考了以往的mmpretrain版本，我将以往能运行的版本进行了代码保存，并与报错代码进行比对，直到修改了data_prefix后才修正，因为每一位开发者的机型与环境都不一定完全相同，您可以根据以往的版本作为参考，看看能不能修复您的报错，您也可以尝试将我的修正代码改变必要的参数后在您的环境运行，希望能够帮助到您。（When I corrected this problem, I referred to the previous versions of mmpretrain. I saved the code of the previous versions that could run and compared them with the error code. I did not correct it until I modified data_prefix, because the model and environment of each developer may not be exactly the same. You can use the previous version as a reference to see if you can fix your error, you can also try my fix code to change the necessary parameters to run in your environment, I hope it can help you.）

vipzms commented 1 month ago

我也遇到了这个错误，你怎么解决呢？你好，朋友，很高兴能够回复你，我现在将我修正后的内容发给你，希望能够帮助到您。（你好，我的朋友，我很高兴回复你。现在我会把更正后的内容发给你，希望能帮到你。

auto_scale_lr = dict(base_batch_size=256)
data_preprocessor = dict(
    mean=[
        123.675,
        116.28,
        103.53,
    ],
    num_classes=7,
    std=[
        58.395,
        57.12,
        57.375,
    ],
    to_rgb=True)
dataset_type = 'ImageNet'
default_hooks = dict(
    checkpoint=dict(interval=1, type='CheckpointHook'),
    logger=dict(interval=100, type='LoggerHook'),
    param_scheduler=dict(type='ParamSchedulerHook'),
    sampler_seed=dict(type='DistSamplerSeedHook'),
    timer=dict(type='IterTimerHook'),
    visualization=dict(enable=False, type='VisualizationHook'))
default_scope = 'mmpretrain'
env_cfg = dict(
    cudnn_benchmark=False,
    dist_cfg=dict(backend='nccl'),
    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
launcher = 'none'
load_from = 'D:\\epoch_50.pth'
log_level = 'INFO'
model = dict(
    backbone=dict(
        depth=50,
        num_stages=4,
        out_indices=(3, ),
        style='pytorch',
        type='ResNet'),
    head=dict(
        in_channels=2048,
        loss=dict(loss_weight=1.0, type='CrossEntropyLoss'),
        num_classes=7,
        topk=(
            1,
            5,
        ),
        type='LinearClsHead'),
    neck=dict(type='GlobalAveragePooling'),
    type='ImageClassifier')
optim_wrapper = dict(
    optimizer=dict(lr=0.1, momentum=0.9, type='SGD', weight_decay=0.0001))
param_scheduler = dict(
    by_epoch=True,
    gamma=0.1,
    milestones=[
        100,
        200,
        300,
        400,
        500,
    ],
    type='MultiStepLR')
randomness = dict(deterministic=False, seed=None)
resume = False
test_cfg = dict()
test_dataloader = dict(
    batch_size=64,
    collate_fn=dict(type='default_collate'),
    dataset=dict(
        data_prefix='test',
        data_root='D:\\data',
        pipeline=[
            dict(type='LoadImageFromFile'),
            dict(edge='short', scale=256, type='ResizeEdge'),
            dict(crop_size=224, type='CenterCrop'),
            dict(type='PackInputs'),
        ],
        type='CustomDataset'),
    num_workers=8,
    persistent_workers=True,
    pin_memory=True,
    sampler=dict(shuffle=False, type='DefaultSampler'))
test_evaluator = [
    dict(topk=(
        1,
        5,
    ), type='Accuracy'),
    dict(
        items=[
            'precision',
            'recall',
            'f1-score',
        ], type='SingleLabelMetric'),
]
test_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(edge='short', scale=256, type='ResizeEdge'),
    dict(crop_size=224, type='CenterCrop'),
    dict(type='PackInputs'),
]
train_cfg = dict(by_epoch=True, max_epochs=500, val_interval=1)
train_dataloader = dict(
    batch_size=64,
    collate_fn=dict(type='default_collate'),
    dataset=dict(
        data_prefix='train',
        data_root='D:\\data',
        pipeline=[
            dict(type='LoadImageFromFile'),
            dict(scale=224, type='RandomResizedCrop'),
            dict(direction='horizontal', prob=0.5, type='RandomFlip'),
            dict(type='PackInputs'),
        ],
        type='CustomDataset'),
    num_workers=8,
    persistent_workers=True,
    pin_memory=True,
    sampler=dict(shuffle=True, type='DefaultSampler'))
train_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(scale=224, type='RandomResizedCrop'),
    dict(direction='horizontal', prob=0.5, type='RandomFlip'),
    dict(type='PackInputs'),
]
val_cfg = dict()
val_dataloader = dict(
    batch_size=64,
    collate_fn=dict(type='default_collate'),
    dataset=dict(
        data_prefix='val',
        data_root='D:\\data',
        pipeline=[
            dict(type='LoadImageFromFile'),
            dict(edge='short', scale=256, type='ResizeEdge'),
            dict(crop_size=224, type='CenterCrop'),
            dict(type='PackInputs'),
        ],
        type='CustomDataset'),
    num_workers=8,
    persistent_workers=True,
    pin_memory=True,
    sampler=dict(shuffle=False, type='DefaultSampler'))
val_evaluator = [
    dict(topk=(
        1,
        5,
    ), type='Accuracy'),
    dict(
        items=[
            'precision',
            'recall',
            'f1-score',
        ], type='SingleLabelMetric'),
]
vis_backends = [
    dict(type='LocalVisBackend'),
]
visualizer = dict(
    type='UniversalVisualizer', vis_backends=[
        dict(type='LocalVisBackend'),
    ])
work_dir = 'D:\\resnet50_8xb32_in1k_epoch500'

你可以重点关注data_prefix=XXX这一部分，将原有的split=XXX进行替换。（您可以专注于 data_prefix=XXX 部分并替换原来的 split=XXX。希望这能帮助到您。（希望这对你有所帮助。

你这个解决了我的问题非常感谢当使用默认resnet18_8xb32_in1k.py进行训练时正好遇到这个问题

open-mmlab / mmpretrain