Gradient accumulation doesn't work well on mask r cnn.

[kazuma-jp]

Summary

Hello.
Thank you for developing mmdetection. I am enjoying it very much!!!

I have a question.
I set gradient accumulation to mask r cnn.
I used the MMCLASSIFICATION’S doc as a reference for the implementation.
But, following code doesn't work well.
optimizer_config = dict(type="GradientCumulativeOptimizerHook", cumulative_iters=4)
Please tell me the proper implementation of gradient accumulation.

ENV

pytorch==1.11
cuda==113
mmcv-full==1.6.1
mmdet==2.25.1

Error

TypeError: __init__() got an unexpected keyword argument 'type'

Code

# user settings
output = 'PATH'
seed = 42 
lr = 2e-3
img_size = (500, 500)
max_epochs = 10

fp16 = dict(loss_scale=512.)
num_classes = 1
classes = ('sample',)
work_dir=output
load_from = 'mask_rcnn_r50_fpn_mstrain-poly_3x_coco_20210524_201154-21b550bb.pth'

# model settings
model = dict(
    type='MaskRCNN',
    backbone=dict(
        type='ResNet',
        depth=50,
        num_stages=4,
        out_indices=(0, 1, 2, 3),
        frozen_stages=1,
        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,
        num_outs=5),
    rpn_head=dict(
        type='RPNHead',
        in_channels=256,
        feat_channels=256,
        anchor_generator=dict(
            type='AnchorGenerator',
            scales=[8],
            ratios=[0.5, 1.0, 2.0],
            strides=[4, 8, 16, 32, 64]),
        bbox_coder=dict(
            type='DeltaXYWHBBoxCoder',
            target_means=[.0, .0, .0, .0],
            target_stds=[1.0, 1.0, 1.0, 1.0]),
        loss_cls=dict(
            type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0),
        loss_bbox=dict(type='L1Loss', loss_weight=1.0)),
    roi_head=dict(
        type='StandardRoIHead',
        bbox_roi_extractor=dict(
            type='SingleRoIExtractor',
            roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0),
            out_channels=256,
            featmap_strides=[4, 8, 16, 32]),
        bbox_head=dict(
            type='Shared2FCBBoxHead',
            in_channels=256,
            fc_out_channels=1024,
            roi_feat_size=7,
            num_classes=num_classes,
            bbox_coder=dict(
                type='DeltaXYWHBBoxCoder',
                target_means=[0., 0., 0., 0.],
                target_stds=[0.1, 0.1, 0.2, 0.2]),
            reg_class_agnostic=False,
            loss_cls=dict(
                type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0),
            loss_bbox=dict(type='L1Loss', loss_weight=1.0)),
        mask_roi_extractor=dict(
            type='SingleRoIExtractor',
            roi_layer=dict(type='RoIAlign', output_size=14, sampling_ratio=0),
            out_channels=256,
            featmap_strides=[4, 8, 16, 32]),
        mask_head=dict(
            type='FCNMaskHead',
            num_convs=4,
            in_channels=256,
            conv_out_channels=256,
            num_classes=num_classes,
            loss_mask=dict(
                type='CrossEntropyLoss', use_mask=True, loss_weight=1.0))),
    # model training and testing settings
    train_cfg=dict(
        rpn=dict(
            assigner=dict(
                type='MaxIoUAssigner',
                pos_iou_thr=0.7,
                neg_iou_thr=0.3,
                min_pos_iou=0.3,
                match_low_quality=True,
                ignore_iof_thr=-1),
            sampler=dict(
                type='RandomSampler',
                num=256,
                pos_fraction=0.5,
                neg_pos_ub=-1,
                add_gt_as_proposals=False),
            allowed_border=-1,
            pos_weight=-1,
            debug=False),
        rpn_proposal=dict(
            nms_pre=2000,#2000
            max_per_img=1000,#1000
            nms=dict(type='nms', iou_threshold=0.7),
            min_bbox_size=0),
        rcnn=dict(
            assigner=dict(
                type='MaxIoUAssigner',
                pos_iou_thr=0.5,
                neg_iou_thr=0.5,
                min_pos_iou=0.5,
                match_low_quality=True,
                ignore_iof_thr=-1),
            sampler=dict(
                type='RandomSampler',
                num=512,
                pos_fraction=0.25,
                neg_pos_ub=-1,
                add_gt_as_proposals=True),
            mask_size=28,
            pos_weight=-1,
            debug=False)),
    test_cfg=dict(
        rpn=dict(
            nms_pre=2000,#2000
            max_per_img=1000,#1000
            nms=dict(type='nms', iou_threshold=0.7),
            min_bbox_size=0),
        rcnn=dict(
            score_thr=0.05,
            nms=dict(type='nms', iou_threshold=0.5),
            max_per_img=100,
            mask_thr_binary=0.5)))

# dataset settings
dataset_type = 'CocoDataset'
data_root = '../data/'
img_norm_cfg = dict(
    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
train_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(type='LoadAnnotations', with_bbox=True, with_mask=True),
    dict(type='Resize', img_scale=img_size, keep_ratio=True),
    dict(type='RandomFlip', flip_ratio=0.5),
    dict(type='Normalize', **img_norm_cfg),
    dict(type='Pad', size_divisor=32),
    dict(type='DefaultFormatBundle'),
    dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels', 'gt_masks']),
]
test_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(
        type='MultiScaleFlipAug',
        img_scale=img_size,
        flip=False,
        transforms=[
            dict(type='Resize', keep_ratio=True),
            dict(type='RandomFlip'),
            dict(type='Normalize', **img_norm_cfg),
            dict(type='Pad', size_divisor=32),
            dict(type='ImageToTensor', keys=['img']),
            dict(type='Collect', keys=['img']),
        ])
]
data = dict(
    samples_per_gpu=1, 
    workers_per_gpu=0,
    train=dict(
        type=dataset_type,
        ann_file=data_root + 'intermediate/train_dataset.json',
        img_prefix=data_root + 'input/train_imgs',
        classes=classes,
        pipeline=train_pipeline),
    val=dict(
        type=dataset_type,
        ann_file=data_root + 'intermediate/val_dataset.json',
        img_prefix=data_root + 'input/train_imgs',
        classes=classes,
        pipeline=test_pipeline),
    test=dict(
        type=dataset_type,
        ann_file=data_root + 'intermediate/val_dataset.json',
        img_prefix=data_root + 'input/train_imgs',
        classes=classes,
        pipeline=test_pipeline))
evaluation = dict(interval=1,
                    metric=['segm'],
                    save_best='segm_mAP')

# optimizer
optimizer = dict(type='AdamW', 
                lr=lr,
                betas=(0.9, 0.999),
                weight_decay=0.01)
optimizer_config = dict(grad_clip=None,
                        type="GradientCumulativeOptimizerHook", 
                        cumulative_iters=4)


# learning policy
lr_config = dict(
    policy='CosineAnnealing',
    by_epoch=False,
    warmup='linear',
    warmup_iters=100,
    warmup_ratio=1.0 / 10,
    min_lr =1e-6
)
runner = dict(type='EpochBasedRunner', max_epochs=max_epochs)

checkpoint_config = dict(interval=1)
log_config = dict(
    interval=50,
    hooks=[
        dict(type='TextLoggerHook'),
        # dict(type='TensorboardLoggerHook')
    ])
custom_hooks = [dict(type='NumClassCheckHook')]

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

[kazuma-jp]

Please someone help me!!

[i-aki-y]

I am not sure but this issue looks to be discussing a similar issue.
#1379

[kazuma-jp]

Thank you for your reply!
I'll check it!!