自己的数据集转换成coco数据集的格式使用不了MAE训练 #706

huadouxiansheng · 2023-02-27T09:27:42Z

huadouxiansheng
Feb 27, 2023

这是我的配置
mmcls 1.0.0rc3
mmcv 2.0.0rc4
mmdet 3.0.0rc5
mmengine 0.3.2
mmsegmentation 1.0.0rc5
mmselfsup 1.0.0rc3 /root/mmselfsup

在MAE的config配置文件中按照教程修改为

# >>>>>>>>>>>>>>>>>>>>> Start of Changed >>>>>>>>>>>>>>>>>>>>>>>>>
_base_ = [
    '../_base_/models/mae_vit-base-p16.py',
    # '../_base_/datasets/imagenet_mae.py',
    '../_base_/schedules/adamw_coslr-200e_in1k.py',
    '../_base_/default_runtime.py',
]

# custom dataset
dataset_type = 'mmdet.CocoDataset'
data_root = 'data/coco/'
file_client_args = dict(backend='disk')

train_pipeline = [
    dict(type='LoadImageFromFile', file_client_args=file_client_args),
    dict(
        type='RandomResizedCrop',
        size=224,
        scale=(0.2, 1.0),
        backend='pillow',
        interpolation='bicubic'),
    dict(type='RandomFlip', prob=0.5),
    dict(type='PackSelfSupInputs', meta_keys=['img_path'])
]

train_dataloader = dict(
    batch_size=128,
    num_workers=8,
    persistent_workers=True,
    sampler=dict(type='DefaultSampler', shuffle=True),
    collate_fn=dict(type='default_collate'),
    dataset=dict(
        type=dataset_type,
        data_root=data_root,
        ann_file='annotations/instances_train2017.json',
        data_prefix=dict(img='train2017/'),
        pipeline=train_pipeline))

# <<<<<<<<<<<<<<<<<<<<<< End of Changed <<<<<<<<<<<<<<<<<<<<<<<<<<<

# optimizer wrapper
optimizer = dict(
    type='AdamW', lr=1.5e-4 * 4096 / 256, betas=(0.9, 0.95), weight_decay=0.05)
optim_wrapper = dict(
    type='OptimWrapper',
    optimizer=optimizer,
    paramwise_cfg=dict(
        custom_keys={
            'ln': dict(decay_mult=0.0),
            'bias': dict(decay_mult=0.0),
            'pos_embed': dict(decay_mult=0.),
            'mask_token': dict(decay_mult=0.),
            'cls_token': dict(decay_mult=0.)
        }))

# learning rate scheduler
param_scheduler = [
    dict(
        type='LinearLR',
        start_factor=1e-4,
        by_epoch=True,
        begin=0,
        end=40,
        convert_to_iter_based=True),
    dict(
        type='CosineAnnealingLR',
        T_max=360,
        by_epoch=True,
        begin=40,
        end=400,
        convert_to_iter_based=True)
]

# runtime settings
# pre-train for 400 epochs
train_cfg = dict(max_epochs=400)
default_hooks = dict(
    logger=dict(type='LoggerHook', interval=100),
    # only keeps the latest 3 checkpoints
    checkpoint=dict(type='CheckpointHook', interval=1, max_keep_ckpts=3))

# randomness
randomness = dict(seed=0, diff_rank_seed=True)
resume = True

因为用了1个gpu，添加了

_base_ = 'mae_vit-base-p16_8xb512-coslr-400e_incoco.py'

# dataset
data = dict(
    samples_per_gpu=32, 
    workers_per_gpu=2,
    train=dict(
        data_source=dict(
            data_prefix='data/coco/train2017',
            ann_file='data/coco/annotations/instances_train2017.json',
        )
    )
)

# optimizer
optimizer = dict(
    lr=1.5e-4 * 4096 / 256 * (32 / 512 * 8),
)

runner = dict(max_epochs=1)

使用训练
!bash tools/dist_train.sh
configs/selfsup/mae/mae_vit-base-p16_1xb32-coslr-1e_incoco.py
1
--work_dir work_dirs/selfsup/mae/mae_vit-base-p16_1xb32-coslr-1e_incoco/
报错

warnings.warn(
Traceback (most recent call last):
  File "tools/train.py", line 9, in <module>
    from mmselfsup.utils import register_all_modules
  File "/root/mmselfsup/mmselfsup/__init__.py", line 2, in <module>
    import mmcls
  File "/root/miniconda3/lib/python3.8/site-packages/mmcls/__init__.py", line 10, in <module>
    mmcv_version = digit_version(mmcv.__version__)
AttributeError: module 'mmcv' has no attribute '__version__'
ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: 1) local_rank: 0 (pid: 17238) of binary: /root/miniconda3/bin/python
Traceback (most recent call last):
  File "/root/miniconda3/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/root/miniconda3/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launch.py", line 195, in <module>
    main()
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launch.py", line 191, in main
    launch(args)
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launch.py", line 176, in launch
    run(args)
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/run.py", line 753, in run
    elastic_launch(
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
    return launch_agent(self._config, self._entrypoint, list(args))
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
    raise ChildFailedError(
torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
============================================================
tools/train.py FAILED
------------------------------------------------------------
Failures:
  <NO_OTHER_FAILURES>
------------------------------------------------------------
Root Cause (first observed failure):
[0]:
  time      : 2023-02-27_17:22:10
  host      : autodl-container-815011ac52-c32bb38f
  rank      : 0 (local_rank: 0)
  exitcode  : 1 (pid: 17238)
  error_file: <N/A>
  traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html
============================================================

Answered by fangyixiao18

Feb 28, 2023

It is recommended to use 1.0.0rc version, it seems your installation of mmcv met some problem, please clear the mmcv in the environment then install it with the guidance in https://github.com/open-mmlab/mmcv/tree/2.x

View full answer

huadouxiansheng · 2023-02-27T12:42:53Z

huadouxiansheng
Feb 27, 2023
Author

重新安装了以前版本的mmselfsup，配置如下

mmcls 0.24.1
mmcv 1.7.0
mmcv-full 1.6.2
mmdet 2.28.1
mmengine 0.3.2
mmsegmentation 0.30.0
mmselfsup 0.11.0 /root/mmselfsup

配置文件和上面相同，新的报错
/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launch.py:180: FutureWarning: The module torch.distributed.launch is deprecated
and will be removed in future. Use torchrun.
Note that --use_env is set by default in torchrun.
If your script expects --local_rank argument to be set, please
change it to read from os.environ['LOCAL_RANK'] instead. See
https://pytorch.org/docs/stable/distributed.html#launch-utility for
further instructions

  warnings.warn(
/root/miniconda3/lib/python3.8/site-packages/mmcv/__init__.py:20: UserWarning: On January 1, 2023, MMCV will release v2.0.0, in which it will remove components related to the training process and add a data transformation module. In addition, it will rename the package names mmcv to mmcv-lite and mmcv-full to mmcv. See https://github.com/open-mmlab/mmcv/blob/master/docs/en/compatibility.md for more details.
  warnings.warn(
fatal: not a git repository (or any of the parent directories): .git
2023-02-27 20:38:47,675 - mmselfsup - INFO - Environment info:
------------------------------------------------------------
sys.platform: linux
Python: 3.8.10 (default, Jun  4 2021, 15:09:15) [GCC 7.5.0]
CUDA available: True
GPU 0: NVIDIA GeForce RTX 3080
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 11.3, V11.3.109
GCC: gcc (Ubuntu 9.4.0-1ubuntu1~20.04.1) 9.4.0
PyTorch: 1.13.0+cu117
PyTorch compiling details: PyTorch built with:
  - GCC 9.3
  - C++ Version: 201402
  - Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v2.6.0 (Git Hash 52b5f107dd9cf10910aaa19cb47f3abf9b349815)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - LAPACK is enabled (usually provided by MKL)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - CUDA Runtime 11.7
  - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86
  - CuDNN 8.5
  - Magma 2.6.1
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.7, CUDNN_VERSION=8.5.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.13.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, 

TorchVision: 0.14.0+cu117
OpenCV: 4.6.0
MMCV: 1.7.1
MMCV Compiler: GCC 9.3
MMCV CUDA Compiler: 11.7
MMSelfSup: 0.11.0+
------------------------------------------------------------

2023-02-27 20:38:47,828 - mmselfsup - INFO - Distributed training: True
2023-02-27 20:38:47,974 - mmselfsup - INFO - Config:
model = dict(
    type='MAE',
    backbone=dict(type='MAEViT', arch='b', patch_size=16, mask_ratio=0.75),
    neck=dict(
        type='MAEPretrainDecoder',
        patch_size=16,
        in_chans=3,
        embed_dim=768,
        decoder_embed_dim=512,
        decoder_depth=8,
        decoder_num_heads=16,
        mlp_ratio=4.0),
    head=dict(type='MAEPretrainHead', norm_pix=True, patch_size=16))
optimizer = dict(type='AdamW', lr=0.0012, betas=(0.9, 0.95), weight_decay=0.05)
optimizer_config = dict()
lr_config = dict(
    policy='CosineAnnealing',
    min_lr=0.0,
    warmup='linear',
    warmup_iters=40,
    warmup_ratio=0.0001,
    warmup_by_epoch=True)
runner = dict(type='EpochBasedRunner', max_epochs=1)
checkpoint_config = dict(interval=10)
log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook')])
dist_params = dict(backend='nccl')
cudnn_benchmark = True
log_level = 'INFO'
load_from = None
resume_from = None
workflow = [('train', 1)]
persistent_workers = True
opencv_num_threads = 0
mp_start_method = 'fork'
dataset_type = 'mmdet.CocoDataset'
data_root = 'data/coco/'
file_client_args = dict(backend='disk')
train_pipeline = [
    dict(type='LoadImageFromFile', file_client_args=dict(backend='disk')),
    dict(
        type='RandomResizedCrop',
        size=224,
        scale=(0.2, 1.0),
        backend='pillow',
        interpolation='bicubic'),
    dict(type='RandomFlip', prob=0.5),
    dict(type='PackSelfSupInputs', meta_keys=['img_path'])
]
train_dataloader = dict(
    batch_size=128,
    num_workers=8,
    persistent_workers=True,
    sampler=dict(type='DefaultSampler', shuffle=True),
    collate_fn=dict(type='default_collate'),
    dataset=dict(
        type='mmdet.CocoDataset',
        data_root='data/coco/',
        ann_file='annotations/instances_train2017.json',
        data_prefix=dict(img='train2017/'),
        pipeline=[
            dict(
                type='LoadImageFromFile',
                file_client_args=dict(backend='disk')),
            dict(
                type='RandomResizedCrop',
                size=224,
                scale=(0.2, 1.0),
                backend='pillow',
                interpolation='bicubic'),
            dict(type='RandomFlip', prob=0.5),
            dict(type='PackSelfSupInputs', meta_keys=['img_path'])
        ]))
optim_wrapper = dict(
    type='OptimWrapper',
    optimizer=dict(
        type='AdamW', lr=0.0024, betas=(0.9, 0.95), weight_decay=0.05),
    paramwise_cfg=dict(
        custom_keys=dict(
            ln=dict(decay_mult=0.0),
            bias=dict(decay_mult=0.0),
            pos_embed=dict(decay_mult=0.0),
            mask_token=dict(decay_mult=0.0),
            cls_token=dict(decay_mult=0.0))))
param_scheduler = [
    dict(
        type='LinearLR',
        start_factor=0.0001,
        by_epoch=True,
        begin=0,
        end=40,
        convert_to_iter_based=True),
    dict(
        type='CosineAnnealingLR',
        T_max=360,
        by_epoch=True,
        begin=40,
        end=400,
        convert_to_iter_based=True)
]
train_cfg = dict(max_epochs=400)
default_hooks = dict(
    logger=dict(type='LoggerHook', interval=100),
    checkpoint=dict(type='CheckpointHook', interval=1, max_keep_ckpts=3))
randomness = dict(seed=0, diff_rank_seed=True)
resume = True
data = dict(
    samples_per_gpu=32,
    workers_per_gpu=2,
    train=dict(
        data_source=dict(
            data_prefix='data/coco/train2017',
            ann_file='data/coco/annotations/train2017.json')))
work_dir = 'work_dirs/selfsup/mae/mae_vit-base-p16_1xb32-coslr-1e_incoco/'
auto_resume = False
gpu_ids = range(0, 1)

2023-02-27 20:38:47,975 - mmselfsup - INFO - Set random seed to 0, deterministic: False
/root/miniconda3/lib/python3.8/site-packages/torch/functional.py:504: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:3190.)
  return _VF.meshgrid(tensors, **kwargs)  # type: ignore[attr-defined]
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.cls_token - torch.Size([1, 1, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.pos_embed - torch.Size([1, 197, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.patch_embed.projection.weight - torch.Size([768, 3, 16, 16]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.patch_embed.projection.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,243 - mmcv - INFO - 
backbone.layers.0.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.0.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.0.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.1.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,244 - mmcv - INFO - 
backbone.layers.2.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.3.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.4.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.5.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,245 - mmcv - INFO - 
backbone.layers.5.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.5.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.6.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.7.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.7.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,246 - mmcv - INFO - 
backbone.layers.7.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.7.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.7.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.7.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.7.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.7.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.7.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.7.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.7.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.7.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.8.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.9.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.9.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.9.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.9.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,247 - mmcv - INFO - 
backbone.layers.9.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.9.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.9.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.9.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.9.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.9.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.9.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.9.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.10.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.11.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.11.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.11.attn.qkv.weight - torch.Size([2304, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.11.attn.qkv.bias - torch.Size([2304]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.11.attn.proj.weight - torch.Size([768, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.11.attn.proj.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.11.ln2.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,248 - mmcv - INFO - 
backbone.layers.11.ln2.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
backbone.layers.11.ffn.layers.0.0.weight - torch.Size([3072, 768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
backbone.layers.11.ffn.layers.0.0.bias - torch.Size([3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
backbone.layers.11.ffn.layers.1.weight - torch.Size([768, 3072]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
backbone.layers.11.ffn.layers.1.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEViT  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
backbone.ln1.weight - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
backbone.ln1.bias - torch.Size([768]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.mask_token - torch.Size([1, 1, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_pos_embed - torch.Size([1, 197, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_embed.weight - torch.Size([512, 768]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_embed.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.ln1.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.ln1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.attn.qkv.weight - torch.Size([1536, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.attn.qkv.bias - torch.Size([1536]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.attn.proj.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.attn.proj.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.ln2.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.ln2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.ffn.layers.0.0.weight - torch.Size([2048, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.ffn.layers.0.0.bias - torch.Size([2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.ffn.layers.1.weight - torch.Size([512, 2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.0.ffn.layers.1.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.1.ln1.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.1.ln1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.1.attn.qkv.weight - torch.Size([1536, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,249 - mmcv - INFO - 
neck.decoder_blocks.1.attn.qkv.bias - torch.Size([1536]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.1.attn.proj.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.1.attn.proj.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.1.ln2.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.1.ln2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.1.ffn.layers.0.0.weight - torch.Size([2048, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.1.ffn.layers.0.0.bias - torch.Size([2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.1.ffn.layers.1.weight - torch.Size([512, 2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.1.ffn.layers.1.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.ln1.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.ln1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.attn.qkv.weight - torch.Size([1536, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.attn.qkv.bias - torch.Size([1536]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.attn.proj.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.attn.proj.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.ln2.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.ln2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.ffn.layers.0.0.weight - torch.Size([2048, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.ffn.layers.0.0.bias - torch.Size([2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.ffn.layers.1.weight - torch.Size([512, 2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.2.ffn.layers.1.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.3.ln1.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.3.ln1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.3.attn.qkv.weight - torch.Size([1536, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.3.attn.qkv.bias - torch.Size([1536]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.3.attn.proj.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,250 - mmcv - INFO - 
neck.decoder_blocks.3.attn.proj.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.3.ln2.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.3.ln2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.3.ffn.layers.0.0.weight - torch.Size([2048, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.3.ffn.layers.0.0.bias - torch.Size([2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.3.ffn.layers.1.weight - torch.Size([512, 2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.3.ffn.layers.1.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.ln1.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.ln1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.attn.qkv.weight - torch.Size([1536, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.attn.qkv.bias - torch.Size([1536]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.attn.proj.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.attn.proj.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.ln2.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.ln2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.ffn.layers.0.0.weight - torch.Size([2048, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.ffn.layers.0.0.bias - torch.Size([2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.ffn.layers.1.weight - torch.Size([512, 2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.4.ffn.layers.1.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.5.ln1.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.5.ln1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.5.attn.qkv.weight - torch.Size([1536, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.5.attn.qkv.bias - torch.Size([1536]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.5.attn.proj.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.5.attn.proj.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.5.ln2.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,251 - mmcv - INFO - 
neck.decoder_blocks.5.ln2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.5.ffn.layers.0.0.weight - torch.Size([2048, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.5.ffn.layers.0.0.bias - torch.Size([2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.5.ffn.layers.1.weight - torch.Size([512, 2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.5.ffn.layers.1.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.ln1.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.ln1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.attn.qkv.weight - torch.Size([1536, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.attn.qkv.bias - torch.Size([1536]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.attn.proj.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.attn.proj.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.ln2.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.ln2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.ffn.layers.0.0.weight - torch.Size([2048, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.ffn.layers.0.0.bias - torch.Size([2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.ffn.layers.1.weight - torch.Size([512, 2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.6.ffn.layers.1.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.ln1.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.ln1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.attn.qkv.weight - torch.Size([1536, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.attn.qkv.bias - torch.Size([1536]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.attn.proj.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.attn.proj.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.ln2.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.ln2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.ffn.layers.0.0.weight - torch.Size([2048, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,252 - mmcv - INFO - 
neck.decoder_blocks.7.ffn.layers.0.0.bias - torch.Size([2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,253 - mmcv - INFO - 
neck.decoder_blocks.7.ffn.layers.1.weight - torch.Size([512, 2048]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,253 - mmcv - INFO - 
neck.decoder_blocks.7.ffn.layers.1.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,253 - mmcv - INFO - 
neck.ln1.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,253 - mmcv - INFO - 
neck.ln1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of MAE  
 
2023-02-27 20:38:50,253 - mmcv - INFO - 
neck.decoder_pred.weight - torch.Size([768, 512]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
2023-02-27 20:38:50,253 - mmcv - INFO - 
neck.decoder_pred.bias - torch.Size([768]): 
Initialized by user-defined `init_weights` in MAEPretrainDecoder  
 
Traceback (most recent call last):
  File "tools/train.py", line 200, in <module>
    main()
  File "tools/train.py", line 182, in main
    datasets = [build_dataset(cfg.data.train)]
  File "/root/mmselfsup/mmselfsup/datasets/builder.py", line 39, in build_dataset
    elif cfg['type'] == 'RepeatDataset':
  File "/root/miniconda3/lib/python3.8/site-packages/mmcv/utils/config.py", line 38, in __missing__
    raise KeyError(name)
KeyError: 'type'
ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: 1) local_rank: 0 (pid: 57177) of binary: /root/miniconda3/bin/python
Traceback (most recent call last):
  File "/root/miniconda3/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/root/miniconda3/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launch.py", line 195, in <module>
    main()
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launch.py", line 191, in main
    launch(args)
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launch.py", line 176, in launch
    run(args)
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/run.py", line 753, in run
    elastic_launch(
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
    return launch_agent(self._config, self._entrypoint, list(args))
  File "/root/miniconda3/lib/python3.8/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
    raise ChildFailedError(
torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
============================================================
tools/train.py FAILED
------------------------------------------------------------
Failures:
  <NO_OTHER_FAILURES>
------------------------------------------------------------
Root Cause (first observed failure):
[0]:
  time      : 2023-02-27_20:38:54
  host      : autodl-container-0bb111be52-079f414c
  rank      : 0 (local_rank: 0)
  exitcode  : 1 (pid: 57177)
  error_file: <N/A>
  traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html
============================================================

0 replies

fangyixiao18 · 2023-02-28T02:49:43Z

fangyixiao18
Feb 28, 2023
Maintainer

Your format in config mae_vit-base-p16_1xb32-coslr-1e_incoco is deprecated. With 1.0.0rc version, you could check the documentation migration

0 replies

fangyixiao18 · 2023-02-28T02:53:17Z

fangyixiao18
Feb 28, 2023
Maintainer

It is recommended to use 1.0.0rc version, it seems your installation of mmcv met some problem, please clear the mmcv in the environment then install it with the guidance in https://github.com/open-mmlab/mmcv/tree/2.x

0 replies

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

自己的数据集转换成coco数据集的格式使用不了MAE训练 #706

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{editor}}'s edit

{{editor}}'s edit

Uh oh!

Replies: 3 comments

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{editor}}'s edit

{{editor}}'s edit

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Select a reply

Uh oh!

自己的数据集转换成coco数据集的格式使用不了MAE训练 #706

Uh oh!

Uh oh!

huadouxiansheng Feb 27, 2023

Replies: 3 comments

Uh oh!

Uh oh!

huadouxiansheng Feb 27, 2023 Author

Uh oh!

fangyixiao18 Feb 28, 2023 Maintainer

Uh oh!

fangyixiao18 Feb 28, 2023 Maintainer

huadouxiansheng
Feb 27, 2023

huadouxiansheng
Feb 27, 2023
Author

fangyixiao18
Feb 28, 2023
Maintainer

fangyixiao18
Feb 28, 2023
Maintainer