_base_ = [
    '/home/liuziyuan/proj/RS-CD/rs-cd-cgwx/configs/_base_/models/changer_mit-b0.py', 
    '/home/liuziyuan/proj/rmcd-kd/configs/common/standard_512x512_200k_cgwx.py']

dataset_type = 'LEVIR_CD_Dataset'
data_root = '/nas/datasets/lzy/RS-ChangeDetection/CGWX'

crop_size = (512, 512)

checkpoint = 'https://download.openmmlab.com/mmsegmentation/v0.5/pretrain/segformer/mit_b0_20220624-7e0fe6dd.pth'  # noqa

model = dict(
    pretrained=checkpoint,
    backbone=dict(
        interaction_cfg=(
            None,
            dict(type='SpatialExchange', p=1/2),
            dict(type='ChannelExchange', p=1/2),
            dict(type='ChannelExchange', p=1/2))
    ),
    decode_head=dict(
        num_classes=2,
        sampler=dict(type='mmseg.OHEMPixelSampler', thresh=0.7, min_kept=100000)),
        # test_cfg=dict(mode='slide', crop_size=crop_size, stride=(crop_size[0]//2, crop_size[1]//2)),
    )

train_pipeline = [
    dict(type='MultiImgLoadImageFromFile'),
    dict(type='MultiImgLoadAnnotations'),
    dict(type='MultiImgRandomRotFlip', rotate_prob=0.5, flip_prob=0.5, degree=(-20, 20)),
    dict(type='MultiImgRandomCrop', crop_size=crop_size, cat_max_ratio=0.75),
    dict(type='MultiImgExchangeTime', prob=0.5),
    dict(
        type='MultiImgPhotoMetricDistortion',
        brightness_delta=10,
        contrast_range=(0.8, 1.2),
        saturation_range=(0.8, 1.2),
        hue_delta=10),
    dict(type='MultiImgPackSegInputs')
]

train_dataloader = dict(
    dataset=dict(pipeline=train_pipeline))

test_pipeline = [
    dict(type='MultiImgLoadImageFromFile'),
    dict(type='MultiImgResize', scale=(512, 512), keep_ratio=True),
    # add loading annotation after ``Resize`` because ground truth
    # does not need to do resize data transform
    dict(type='MultiImgLoadAnnotations'),
    dict(type='MultiImgPackSegInputs')
]


# optimizer
optimizer=dict(
    type='AdamW', lr=0.0001, betas=(0.9, 0.999), weight_decay=0.01)

optim_wrapper = dict(
    _delete_=True,
    type='OptimWrapper',
    optimizer=optimizer,
    paramwise_cfg=dict(
        custom_keys={
            'pos_block': dict(decay_mult=0.),
            'norm': dict(decay_mult=0.),
            'head': dict(lr_mult=10.)
        }))

param_scheduler = [
    dict(
        type='LinearLR', start_factor=1e-6, by_epoch=False, begin=0, end=1000),
    dict(
        type='PolyLR',
        power=1.0,
        begin=1000,
        end=200000,
        eta_min=0.0,
        by_epoch=False,
    )
]
# training schedule for 100k
train_cfg = dict(type='IterBasedTrainLoop', max_iters=200000, val_interval=1000)
val_cfg = dict(type='ValLoop')
test_cfg = dict(type='TestLoop')
default_hooks = dict(
    timer=dict(type='IterTimerHook'),
    logger=dict(type='LoggerHook', interval=50, log_metric_by_epoch=False),
    param_scheduler=dict(type='ParamSchedulerHook'),
    checkpoint=dict(type='CheckpointHook', by_epoch=False, interval=1000,
                    save_best='mIoU'),
    sampler_seed=dict(type='DistSamplerSeedHook'),
    visualization=dict(type='CDVisualizationHook', interval=1, 
                       img_shape=(512, 512, 3)))