ValueError: need at least one array to concatenate

[arymax]

I have searching for this problem two days plz give me some help !

I want using my own dataset to finetune a pretrain model and I have change it to COCO dataset
following is my config
Config:
auto_scale_lr = dict(base_batch_size=64, enable=True)
backend_args = dict(backend='local')
codec = dict(
heatmap_size=(
48,
64,
),
input_size=(
192,
256,
),
sigma=2,
type='UDPHeatmap')
custom_hooks = [
dict(type='SyncBuffersHook'),
]
custom_imports = dict(
allow_failed_imports=False,
imports=[
'mmpose.engine.optim_wrappers.layer_decay_optim_wrapper',
])
data_mode = 'topdown'
data_root = 'data/coco/'
dataset_info = dict(
data_root='/data1/data/vscode-ml/pose detect/確認使用',
dataset_name='pelvictiltDataset',
joint_weights=[
1.5,
1.2,
1.0,
1.0,
1.2,
1.5,
1.3,
],
keypoint_info=dict({
0:
dict(color=[
255,
128,
0,
], id=0, name='head', swap='', type='upper'),
1:
dict(color=[
255,
128,
0,
], id=1, name='ear', swap='', type='upper'),
2:
dict(
color=[
255,
128,
0,
],
id=2,
name='shoulder',
swap='',
type='upper'),
3:
dict(
color=[
0,
255,
0,
], id=3, name='psis', swap='psis', type='lower'),
4:
dict(
color=[
0,
255,
0,
], id=4, name='asis', swap='asis', type='lower'),
5:
dict(color=[
0,
255,
0,
], id=5, name='knee', swap='', type='lower'),
6:
dict(
color=[
0,
0,
255,
],
id=6,
name='midpoint_psis_asis',
swap='',
type='lower')
}),
paper_info=dict(author='max'),
sigmas=[
0.089,
0.087,
0.107,
0.107,
0.087,
0.089,
0.095,
],
skeleton_info=dict({
0:
dict(color=[
255,
128,
0,
], id=0, link=(
'head',
'ear',
)),
1:
dict(color=[
255,
128,
0,
], id=1, link=(
'head',
'shoulder',
)),
2:
dict(color=[
0,
255,
0,
], id=2, link=(
'shoulder',
'psis',
)),
3:
dict(color=[
0,
255,
0,
], id=3, link=(
'psis',
'asis',
)),
4:
dict(color=[
0,
255,
0,
], id=4, link=(
'asis',
'knee',
)),
5:
dict(color=[
0,
0,
255,
], id=5, link=(
'psis',
'midpoint_psis_asis',
)),
6:
dict(color=[
0,
0,
255,
], id=6, link=(
'asis',
'midpoint_psis_asis',
))
}))
dataset_type = 'CocoDataset'
default_hooks = dict(
checkpoint=dict(
interval=10,
max_keep_ckpts=1,
rule='greater',
save_best='coco/AP',
type='CheckpointHook'),
logger=dict(interval=50, type='LoggerHook'),
param_scheduler=dict(type='ParamSchedulerHook'),
sampler_seed=dict(type='DistSamplerSeedHook'),
timer=dict(type='IterTimerHook'),
visualization=dict(enable=False, type='PoseVisualizationHook'))
default_scope = 'mmpose'
env_cfg = dict(
cudnn_benchmark=False,
dist_cfg=dict(backend='nccl'),
mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
load_from = None
log_level = 'INFO'
log_processor = dict(
by_epoch=True, num_digits=6, type='LogProcessor', window_size=50)
model = dict(
backbone=dict(
arch='huge',
drop_path_rate=0.55,
frozen_stages=2,
img_size=(
256,
192,
),
init_cfg=dict(
checkpoint=
'https://download.openmmlab.com/mmpose/v1/pretrained_models/mae_pretrain_vit_huge.pth',
prefix='backbone',
type='Pretrained'),
out_type='featmap',
patch_cfg=dict(padding=2),
patch_size=16,
qkv_bias=True,
type='mmpretrain.VisionTransformer',
with_cls_token=False),
data_preprocessor=dict(
bgr_to_rgb=True,
mean=[
123.675,
116.28,
103.53,
],
std=[
58.395,
57.12,
57.375,
],
type='PoseDataPreprocessor'),
head=dict(
decoder=dict(
heatmap_size=(
48,
64,
),
input_size=(
192,
256,
),
sigma=2,
type='UDPHeatmap'),
deconv_kernel_sizes=(
4,
4,
),
deconv_out_channels=(
256,
256,
),
in_channels=1280,
loss=dict(type='KeypointMSELoss', use_target_weight=True),
out_channels=7,
type='HeatmapHead'),
test_cfg=dict(flip_mode='heatmap', flip_test=True, shift_heatmap=False),
type='TopdownPoseEstimator')
optim_wrapper = dict(
clip_grad=dict(max_norm=1.0, norm_type=2),
constructor='LayerDecayOptimWrapperConstructor',
optimizer=dict(
betas=(
0.9,
0.999,
), lr=0.0005, type='AdamW', weight_decay=0.1),
paramwise_cfg=dict(
custom_keys=dict(
bias=dict(decay_multi=0.0),
norm=dict(decay_mult=0.0),
pos_embed=dict(decay_mult=0.0),
relative_position_bias_table=dict(decay_mult=0.0)),
layer_decay_rate=0.85,
num_layers=32))
param_scheduler = [
dict(
begin=0, by_epoch=False, end=500, start_factor=0.001, type='LinearLR'),
dict(
begin=0,
by_epoch=True,
end=210,
gamma=0.1,
milestones=[
170,
200,
],
type='MultiStepLR'),
]
resume = False
test_cfg = dict()
test_dataloader = dict(
batch_size=8,
dataset=dict(
ann_file='/data1/data/vscode-ml/pose detect/test_output.json',
bbox_file='/data1/data/vscode-ml/pose detect/bbox_output.json',
data_mode='topdown',
data_prefix=dict(
img='/data1/data/vscode-ml/pose detect/確認使用/test_images'),
data_root='/data1/data/vscode-ml/pose detect/確認使用',
metainfo=dict(
from_file=
'/data1/data/vscode-ml/pose detect/mmpose/configs/base/datasets/pelvictilt-custom.py'
),
pipeline=[
dict(type='LoadImage'),
dict(type='GetBBoxCenterScale'),
dict(input_size=(
192,
256,
), type='TopdownAffine', use_udp=True),
dict(type='PackPoseInputs'),
],
test_mode=True,
type='CocoDataset'),
drop_last=False,
num_workers=4,
persistent_workers=True,
sampler=dict(round_up=False, shuffle=False, type='DefaultSampler'))
test_evaluator = dict(
ann_file='/data1/data/vscode-ml/pose detect/test_output.json',
type='CocoMetric')
train_cfg = dict(by_epoch=True, max_epochs=210, val_interval=10)
train_dataloader = dict(
batch_size=16,
dataset=dict(
ann_file='/data1/data/vscode-ml/pose detect/train_output.json',
data_mode='topdown',
data_prefix=dict(
img='/data1/data/vscode-ml/pose detect/確認使用/train_images'),
data_root='/data1/data/vscode-ml/pose detect/確認使用/',
metainfo=dict(
from_file=
'/data1/data/vscode-ml/pose detect/mmpose/configs/base/datasets/pelvictilt-custom.py'
),
pipeline=[
dict(type='LoadImage'),
dict(type='GetBBoxCenterScale'),
dict(direction='horizontal', type='RandomFlip'),
dict(type='RandomHalfBody'),
dict(type='RandomBBoxTransform'),
dict(input_size=(
192,
256,
), type='TopdownAffine', use_udp=True),
dict(
encoder=dict(
heatmap_size=(
48,
64,
),
input_size=(
192,
256,
),
sigma=2,
type='UDPHeatmap'),
type='GenerateTarget'),
dict(type='PackPoseInputs'),
],
type='CocoDataset'),
num_workers=4,
persistent_workers=True,
sampler=dict(shuffle=True, type='DefaultSampler'))
train_pipeline = [
dict(type='LoadImage'),
dict(type='GetBBoxCenterScale'),
dict(direction='horizontal', type='RandomFlip'),
dict(type='RandomHalfBody'),
dict(type='RandomBBoxTransform'),
dict(input_size=(
192,
256,
), type='TopdownAffine', use_udp=True),
dict(
encoder=dict(
heatmap_size=(
48,
64,
),
input_size=(
192,
256,
),
sigma=2,
type='UDPHeatmap'),
type='GenerateTarget'),
dict(type='PackPoseInputs'),
]
val_cfg = dict()
val_dataloader = dict(
batch_size=8,
dataset=dict(
ann_file='/data1/data/vscode-ml/pose detect/val_output.json',
bbox_file='/data1/data/vscode-ml/pose detect/bbox_output.json',
data_mode='topdown',
data_prefix=dict(
img='/data1/data/vscode-ml/pose detect/確認使用/val_images'),
data_root='/data1/data/vscode-ml/pose detect/確認使用',
metainfo=dict(
from_file=
'/data1/data/vscode-ml/pose detect/mmpose/configs/base/datasets/pelvictilt-custom.py'
),
pipeline=[
dict(type='LoadImage'),
dict(type='GetBBoxCenterScale'),
dict(input_size=(
192,
256,
), type='TopdownAffine', use_udp=True),
dict(type='PackPoseInputs'),
],
test_mode=True,
type='CocoDataset'),
drop_last=False,
num_workers=4,
persistent_workers=True,
sampler=dict(round_up=False, shuffle=False, type='DefaultSampler'))
val_evaluator = dict(
ann_file='/data1/data/vscode-ml/pose detect/val_output.json',
type='CocoMetric')
val_pipeline = [
dict(type='LoadImage'),
dict(type='GetBBoxCenterScale'),
dict(input_size=(
192,
256,
), type='TopdownAffine', use_udp=True),
dict(type='PackPoseInputs'),
]
vis_backends = [
dict(type='LocalVisBackend'),
]
visualizer = dict(
name='visualizer',
type='PoseLocalVisualizer',
vis_backends=[
dict(type='LocalVisBackend'),
])
am I do anything wrong on this config?