复现出来的结果比较低，请求帮助。

[Rsodlei]

作者你好，我在本地复现了PWOOD-HBox的实验，数据切分按照SOOD进行，在10%上的结果是0.524376，在20%的结果是0.579017，在30%的结果是0.593。我在您提供的配置文件中只修改了数据位置。在30%上结果相差较大，请问这种波动是正常现象吗。

angle_version = 'le90'

import torchvision.transforms as transforms
from copy import deepcopy

model settings

detector = dict(
type='SemiRotatedFCOS',
backbone=dict(
type='ResNet',
depth=50,
num_stages=4,
out_indices=(0, 1, 2, 3),
frozen_stages=1,
zero_init_residual=False,
norm_cfg=dict(type='BN', requires_grad=True),
norm_eval=True,
style='pytorch',
init_cfg=dict(type='Pretrained', checkpoint='torchvision://resnet50')),
neck=dict(
type='FPN',
in_channels=[256, 512, 1024, 2048],
out_channels=256,
start_level=1,
add_extra_convs='on_output', # use P5
num_outs=5,
relu_before_extra_convs=True),
bbox_head=dict(
type='SemiRotatedFCOSHeadH2RV2MCL',
num_classes=16,
in_channels=256,
stacked_convs=4,
feat_channels=256,
strides=[8, 16, 32, 64, 128],
center_sampling=True,
center_sample_radius=1.5,
norm_on_bbox=True,
centerness_on_reg=True,
#separate_angle=False,
square_cls=[1, 9, 11],
resize_cls=[1],
scale_angle=False,
bbox_coder=dict(
type='DistanceAnglePointCoder', angle_version=angle_version),
# sup_loss-------------
loss_cls=dict(
type='FocalLoss',
use_sigmoid=True,
gamma=2.0,
alpha=0.25,
loss_weight=1.0),
loss_bbox=dict(type='IoULoss', loss_weight=1.0),
loss_centerness=dict(
type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0),
# aug_loss----------------
loss_ss_symmetry=dict(
type='SmoothL1Loss', loss_weight=0.2, beta=0.1)),

# training and testing settings
train_cfg=None,
test_cfg=dict(
    nms_pre=2000,
    min_bbox_size=0,
    score_thr=0.05,
    nms=dict(iou_thr=0.1),
    max_per_img=2000))

model = dict(
type="H2RV2MCLTeacher",
model=detector,
semi_loss=dict(type='SemiGMMLoss', cls_channels=16),
train_cfg=dict(
iter_count=0,
burn_in_steps=12800,
sup_weight=1.0,
unsup_weight=1.0,
weight_suppress="exp",
logit_specific_weights=dict(),
cls_channels=16
),
test_cfg=dict(inference_on="teacher"),
)

img_norm_cfg = dict(
mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
common_pipeline = [
dict(type='Normalize', **img_norm_cfg),
dict(type='Pad', size_divisor=32),
dict(type='DefaultFormatBundle'),
dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels'],
meta_keys=('filename', 'ori_filename', 'ori_shape',
'img_shape', 'pad_shape', 'scale_factor', 'flip', 'flip_direction', 'img_norm_cfg', 'tag')

     )

]
strong_pipeline = [
dict(type='DTToPILImage'),
dict(type='DTRandomApply', operations=[transforms.ColorJitter(0.4, 0.4, 0.4, 0.1)], p=0.8),
dict(type='DTRandomGrayscale', p=0.2),
dict(type='DTRandomApply', operations=[
dict(type='DTGaussianBlur', rad_range=[0.1, 2.0])
]),
# dict(type='DTRandCrop'),
dict(type='DTToNumpy'),
dict(type="ExtraAttrs", tag="unsup_strong"),
]
weak_pipeline = [
# dict(type='LoadImageFromFile'),
# dict(type='LoadAnnotations', with_bbox=True),
dict(type='RResize', img_scale=(1024, 1024)),
dict(
type='RRandomFlip',
flip_ratio=[0.25, 0.25, 0.25],
direction=['horizontal', 'vertical', 'diagonal'],
version=angle_version),
dict(type="ExtraAttrs", tag="unsup_weak"),
]
unsup_pipeline = [
dict(type="LoadImageFromFile"),
# dict(type="LoadAnnotations", with_bbox=True),
# generate fake labels for data format compatibility
dict(type="LoadEmptyAnnotations", with_bbox=True),
dict(type="STMultiBranch", unsup_strong=deepcopy(strong_pipeline), unsup_weak=deepcopy(weak_pipeline),
common_pipeline=common_pipeline, is_seq=True),
]
sup_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='LoadAnnotations', with_bbox=True),
dict(type='RResize', img_scale=(1024, 1024)),
dict(
type='RRandomFlip',
flip_ratio=[0.25, 0.25, 0.25],
direction=['horizontal', 'vertical', 'diagonal'],
version=angle_version),
dict(type="ExtraAttrs", tag="sup_weak"),
dict(type='Normalize', **img_norm_cfg),
dict(type='Pad', size_divisor=32),
dict(type='DefaultFormatBundle'),
dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels'],
meta_keys=('filename', 'ori_filename', 'ori_shape',
'img_shape', 'pad_shape', 'scale_factor', 'flip',
'flip_direction', 'img_norm_cfg', 'tag')
)
]
test_pipeline = [
dict(type='LoadImageFromFile'),
dict(
type='MultiScaleFlipAug',
img_scale=(1024, 1024),
flip=False,
transforms=[
dict(type='RResize'),
dict(type='Normalize', **img_norm_cfg),
dict(type='Pad', size_divisor=32),
dict(type='DefaultFormatBundle'),
dict(type='Collect', keys=['img'])
])
]

dataset_type = 'DOTAv15WSOODDataset'
classes = ('plane', 'baseball-diamond', 'bridge', 'ground-track-field',
'small-vehicle', 'large-vehicle', 'ship', 'tennis-court',
'basketball-court', 'storage-tank', 'soccer-ball-field',
'roundabout', 'harbor', 'swimming-pool', 'helicopter',
'container-crane')
data = dict(
samples_per_gpu=3,
workers_per_gpu=5,
train=dict(
type="SemiDataset",
sup=dict(
type=dataset_type,
pipeline=sup_pipeline,
ann_file="/home/l/DOTAv1.5/split_ss_dota1_5/train_30p_labeled/annfiles/",
img_prefix="/home/l/DOTAv1.5/split_ss_dota1_5/train_30p_labeled/images/",
version=angle_version,
classes=classes
),
unsup=dict(
type=dataset_type,
pipeline=unsup_pipeline,
ann_file="/home/l/DOTAv1.5/split_ss_dota1_5/train_30p_unlabeled/empty_annfiles/",
img_prefix="/home/l/DOTAv1.5/split_ss_dota1_5/train_30p_unlabeled/images/",
version=angle_version,
classes=classes,
filter_empty_gt=False
),
),
val=dict(
type=dataset_type,
pipeline=test_pipeline,
img_prefix="/home/l/DOTAv1.5/split_ss_dota1_5/val/images/",
ann_file='/home/l/DOTAv1.5/split_ss_dota1_5/val/annfiles/',
version=angle_version,
classes=classes
),
test=dict(
type=dataset_type,
pipeline=test_pipeline,
img_prefix="/home/l/DOTAv1.5/split_ss_dota1_5/test/images/",
ann_file='/home/l/DOTAv1.5/split_ss_dota1_5/test/annfiles/',
version=angle_version,
classes=classes

),
sampler=dict(
    train=dict(
        type="MultiSourceSampler",
        sample_ratio=[2, 1],
        seed=42
    )
),

)

custom_hooks = [
dict(type="NumClassCheckHook"),
# dict(type="WeightSummary"),
dict(type="MeanTeacher", momentum=0.9996, interval=1, start_steps=3200),
]

evaluation

evaluation = dict(type="SubModulesDistEvalHook", interval=3200, metric='mAP',
save_best='mAP')

optimizer

#optimizer = dict(type='SGD', lr=0.0025, momentum=0.9, weight_decay=0.0001)
optimizer = dict(
#delete=True,
type='AdamW',
lr=0.0001,
betas=(0.9, 0.999),
weight_decay=0.05)
optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))

learning policy

lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=1.0 / 3,
step=[120000, 160000])

120k iters is enough for DOTA

runner = dict(type="IterBasedRunner", max_iters=180000)
checkpoint_config = dict(by_epoch=False, interval=3200, max_keep_ckpts=1)

Default: disable fp16 training

fp16 = dict(loss_scale="dynamic")

log_config = dict(
delete=True,
interval=50,
hooks=[
dict(type="TextLoggerHook"),
# dict(
# type="WandbLoggerHook",
# init_kwargs=dict(
# project="rotated_DenseTeacher_10percent",
# name="default_bce4cls",
# ),
# by_epoch=False,
# ),
],
)

dist_params = dict(backend='nccl')
log_level = 'INFO'
load_from = None
resume_from = None
workflow = [('train', 1)] # mode, iters

disable opencv multithreading to avoid system being overloaded

opencv_num_threads = 0

set multi-process start method as fork to speed up the training

mp_start_method = 'fork'

[123sio]

一般来说，在 2 个点左右的波动我们认为是可以接受的。建议在相同配置条件下多次重复实验，以观察结果的稳定性

[Rsodlei]

好的！谢谢您的回复！