update codes

mmdet/configs/dino/dino_4scale_r50_8xb2_24e_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+from mmengine.config import read_base
+from mmengine.runner.loops import EpochBasedTrainLoop
+
+with read_base():
+    from .dino_4scale_r50_8xb2_12e_coco import *
+
+max_epochs = 24
+train_cfg.update(
+    dict(type=EpochBasedTrainLoop, max_epochs=max_epochs, val_interval=1))
+
+param_scheduler[0].update(dict(milestones=[20]))

mmdet/configs/dino/dino_4scale_r50_8xb2_36e_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+from mmengine.config import read_base
+from mmengine.runner.loops import EpochBasedTrainLoop
+
+with read_base():
+    from .dino_4scale_r50_8xb2_12e_coco import *
+
+max_epochs = 36
+train_cfg.update(
+    dict(type=EpochBasedTrainLoop, max_epochs=max_epochs, val_interval=1))
+
+param_scheduler[0].update(dict(milestones=[30]))

mmdet/configs/dino/dino_4scale_r50_improved_8xb2_12e_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+from mmengine.config import read_base
+
+with read_base():
+    from .dino_4scale_r50_8xb2_12e_coco import *
+
+# from deformable detr hyper
+model.update(
+    dict(
+        backbone=dict(frozen_stages=-1),
+        bbox_head=dict(loss_cls=dict(loss_weight=2.0)),
+        positional_encoding=dict(offset=-0.5, temperature=10000),
+        dn_cfg=dict(group_cfg=dict(num_dn_queries=300))))
+
+# optimizer
+optim_wrapper.update(
+    dict(
+        optimizer=dict(lr=0.0002),
+        paramwise_cfg=dict(
+            custom_keys={
+                'backbone': dict(lr_mult=0.1),
+                'sampling_offsets': dict(lr_mult=0.1),
+                'reference_points': dict(lr_mult=0.1)
+            })))

mmdet/configs/dino/dino_5scale_swin_l_8xb2_12e_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+from mmengine.config import read_base
+from mmengine.model.weight_init import PretrainedInit
+
+from mmdet.models import SwinTransformer
+
+with read_base():
+    from .dino_4scale_r50_8xb2_12e_coco import *
+
+pretrained = 'https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_large_patch4_window12_384_22k.pth'  # noqa
+num_levels = 5
+model.merge(
+    dict(
+        num_feature_levels=num_levels,
+        backbone=dict(
+            _delete_=True,
+            type=SwinTransformer,
+            pretrain_img_size=384,
+            embed_dims=192,
+            depths=[2, 2, 18, 2],
+            num_heads=[6, 12, 24, 48],
+            window_size=12,
+            mlp_ratio=4,
+            qkv_bias=True,
+            qk_scale=None,
+            drop_rate=0.,
+            attn_drop_rate=0.,
+            drop_path_rate=0.2,
+            patch_norm=True,
+            out_indices=(0, 1, 2, 3),
+            # Please only add indices that would be used
+            # in FPN, otherwise some parameter will not be used
+            with_cp=True,
+            convert_weights=True,
+            init_cfg=dict(type=PretrainedInit, checkpoint=pretrained)),
+        neck=dict(in_channels=[192, 384, 768, 1536], num_outs=num_levels),
+        encoder=dict(
+            layer_cfg=dict(self_attn_cfg=dict(num_levels=num_levels))),
+        decoder=dict(
+            layer_cfg=dict(cross_attn_cfg=dict(num_levels=num_levels)))))

mmdet/configs/dino/dino_5scale_swin_l_8xb2_36e_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+from mmengine.config import read_base
+from mmengine.runner.loops import EpochBasedTrainLoop
+
+with read_base():
+    from .dino_5scale_swin_l_8xb2_12e_coco import *
+
+max_epochs = 36
+train_cfg.update(
+    dict(type=EpochBasedTrainLoop, max_epochs=max_epochs, val_interval=1))
+
+param_scheduler[0].update(dict(milestones=[27, 33]))

mmdet/configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .._base_.datasets.coco_detection import *
+    from .._base_.default_runtime import *
+    from .._base_.models.faster_rcnn_r50_fpn import *
+    from .._base_.schedules.schedule_1x import *

mmdet/configs/mask_rcnn/mask_rcnn_r101_caffe_fpn_1x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r50_fpn_poly_1x_coco import *
+
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    backbone=dict(
+        depth=101,
+        init_cfg=dict(
+            type=PretrainedInit,
+            checkpoint='open-mmlab://detectron2/resnet101_caffe')))

mmdet/configs/mask_rcnn/mask_rcnn_r101_caffe_fpn_ms_poly_3x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from ..common.ms_poly_3x_coco_instance import *
+    from .._base_.models.mask_rcnn_r50_fpn import *
+
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    # use caffe img_norm
+    data_preprocessor=dict(
+        mean=[103.530, 116.280, 123.675],
+        std=[1.0, 1.0, 1.0],
+        bgr_to_rgb=False),
+    backbone=dict(
+        depth=101,
+        norm_cfg=dict(requires_grad=False),
+        norm_eval=True,
+        style='caffe',
+        init_cfg=dict(
+            type=PretrainedInit,
+            checkpoint='open-mmlab://detectron2/resnet101_caffe')))

mmdet/configs/mask_rcnn/mask_rcnn_r101_fpn_1x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .._base_.models.mask_rcnn_r50_fpn import *
+
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    backbone=dict(
+        depth=101,
+        init_cfg=dict(
+            type=PretrainedInit, checkpoint='torchvision://resnet101')))

mmdet/configs/mask_rcnn/mask_rcnn_r101_fpn_2x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r50_fpn_2x_coco import *
+
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    backbone=dict(
+        depth=101,
+        init_cfg=dict(
+            type=PretrainedInit, checkpoint='torchvision://resnet101')))

mmdet/configs/mask_rcnn/mask_rcnn_r101_fpn_8xb8_amp_lsj_200e_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r18_fpn_8xb8_amp_lsj_200e_coco import *
+
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    backbone=dict(
+        depth=101,
+        init_cfg=dict(
+            type=PretrainedInit, checkpoint='torchvision://resnet101')))

mmdet/configs/mask_rcnn/mask_rcnn_r101_fpn_ms_poly_3x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from ..common.ms_poly_3x_coco_instance import *
+    from .._base_.models.mask_rcnn_r50_fpn import *
+
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    backbone=dict(
+        depth=101,
+        init_cfg=dict(
+            type=PretrainedInit, checkpoint='torchvision://resnet101')))

mmdet/configs/mask_rcnn/mask_rcnn_r18_fpn_8xb8_amp_lsj_200e_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r50_fpn_8xb8_amp_lsj_200e_coco import *
+
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    backbone=dict(
+        depth=18,
+        init_cfg=dict(
+            type=PretrainedInit, checkpoint='torchvision://resnet18')),
+    neck=dict(in_channels=[64, 128, 256, 512]))

mmdet/configs/mask_rcnn/mask_rcnn_r50_caffe_c4_1x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .._base_.datasets.coco_instance import *
+    from .._base_.default_runtime import *
+    from .._base_.models.mask_rcnn_r50_caffe_c4 import *
+    from .._base_.schedules.schedule_1x import *

mmdet/configs/mask_rcnn/mask_rcnn_r50_caffe_fpn_1x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r50_fpn_1x_coco import *
+
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    # use caffe img_norm
+    data_preprocessor=dict(
+        mean=[103.530, 116.280, 123.675],
+        std=[1.0, 1.0, 1.0],
+        bgr_to_rgb=False),
+    backbone=dict(
+        norm_cfg=dict(requires_grad=False),
+        style='caffe',
+        init_cfg=dict(
+            type=PretrainedInit,
+            checkpoint='open-mmlab://detectron2/resnet50_caffe')))

mmdet/configs/mask_rcnn/mask_rcnn_r50_caffe_fpn_ms_1x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r50_fpn_1x_coco import *
+
+from mmcv.transforms import RandomChoiceResize
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    # use caffe img_norm
+    data_preprocessor=dict(
+        mean=[103.530, 116.280, 123.675],
+        std=[1.0, 1.0, 1.0],
+        bgr_to_rgb=False),
+    backbone=dict(
+        norm_cfg=dict(requires_grad=False),
+        style='caffe',
+        init_cfg=dict(
+            type=PretrainedInit,
+            checkpoint='open-mmlab://detectron2/resnet50_caffe')))
+
+train_pipeline = [
+    dict(type=LoadImageFromFile, backend_args={{_base_.backend_args}}),
+    dict(type=LoadAnnotations, with_bbox=True, with_mask=True),
+    dict(
+        type=RandomChoiceResize,
+        scales=[(1333, 640), (1333, 672), (1333, 704), (1333, 736),
+                (1333, 768), (1333, 800)],
+        keep_ratio=True),
+    dict(type=RandomFlip, prob=0.5),
+    dict(type=PackDetInputs),
+]
+
+train_dataloader.update(dict(dataset=dict(pipeline=train_pipeline)))

mmdet/configs/mask_rcnn/mask_rcnn_r50_caffe_fpn_ms_poly_1x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r50_fpn_1x_coco import *
+
+from mmcv.transforms import RandomChoiceResize
+from mmengine.model.weight_init import PretrainedInit
+
+model = dict(
+    # use caffe img_norm
+    data_preprocessor=dict(
+        mean=[103.530, 116.280, 123.675],
+        std=[1.0, 1.0, 1.0],
+        bgr_to_rgb=False),
+    backbone=dict(
+        norm_cfg=dict(requires_grad=False),
+        style='caffe',
+        init_cfg=dict(
+            type=PretrainedInit,
+            checkpoint='open-mmlab://detectron2/resnet50_caffe')))
+train_pipeline = [
+    dict(type=LoadImageFromFile, backend_args={{_base_.backend_args}}),
+    dict(
+        type=LoadAnnotations, with_bbox=True, with_mask=True, poly2mask=False),
+    dict(
+        type=RandomChoiceResize,
+        scales=[(1333, 640), (1333, 672), (1333, 704), (1333, 736),
+                (1333, 768), (1333, 800)],
+        keep_ratio=True),
+    dict(type=RandomFlip, prob=0.5),
+    dict(type=PackDetInputs)
+]
+
+train_dataloader.update(dict(dataset=dict(pipeline=train_pipeline)))

mmdet/configs/mask_rcnn/mask_rcnn_r50_caffe_fpn_ms_poly_2x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r50_caffe_fpn_ms_poly_1x_coco import *
+
+train_cfg = dict(max_epochs=24)
+# learning rate
+param_scheduler = [
+    dict(type=LinearLR, start_factor=0.001, by_epoch=False, begin=0, end=500),
+    dict(
+        type=MultiStepLR,
+        begin=0,
+        end=24,
+        by_epoch=True,
+        milestones=[16, 22],
+        gamma=0.1)
+]

mmdet/configs/mask_rcnn/mask_rcnn_r50_caffe_fpn_ms_poly_3x_coco.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r50_caffe_fpn_ms_poly_1x_coco import *
+
+train_cfg = dict(max_epochs=36)
+# learning rate
+param_scheduler = [
+    dict(type=LinearLR, start_factor=0.001, by_epoch=False, begin=0, end=500),
+    dict(
+        type=MultiStepLR,
+        begin=0,
+        end=24,
+        by_epoch=True,
+        milestones=[28, 34],
+        gamma=0.1)
+]

mmdet/configs/mask_rcnn/mask_rcnn_r50_caffe_fpn_poly_1x_coco_v1.py

+# Copyright (c) OpenMMLab. All rights reserved.
+
+# Please refer to https://mmengine.readthedocs.io/en/latest/advanced_tutorials/config.html#a-pure-python-style-configuration-file-beta for more details. # noqa
+# mmcv >= 2.0.1
+# mmengine >= 0.8.0
+
+from mmengine.config import read_base
+
+with read_base():
+    from .mask_rcnn_r50_fpn_1x_coco import *
+
+from mmengine.model.weight_init import PretrainedInit
+
+from mmdet.models.losses import SmoothL1Loss
+
+model = dict(
+    # use caffe img_norm
+    data_preprocessor=dict(
+        mean=[103.530, 116.280, 123.675],
+        std=[1.0, 1.0, 1.0],
+        bgr_to_rgb=False),
+    backbone=dict(
+        norm_cfg=dict(requires_grad=False),
+        style='caffe',
+        init_cfg=dict(
+            type=PretrainedInit,
+            checkpoint='open-mmlab://detectron2/resnet50_caffe')),
+    rpn_head=dict(
+        loss_bbox=dict(type=SmoothL1Loss, beta=1.0 / 9.0, loss_weight=1.0)),
+    roi_head=dict(
+        bbox_roi_extractor=dict(
+            roi_layer=dict(
+                type=RoIAlign, output_size=7, sampling_ratio=2,
+                aligned=False)),
+        bbox_head=dict(
+            loss_bbox=dict(type=SmoothL1Loss, beta=1.0, loss_weight=1.0)),
+        mask_roi_extractor=dict(
+            roi_layer=dict(
+                type=RoIAlign, output_size=14, sampling_ratio=2,
+                aligned=False))))