model = dict(
    type="TadTR",  # Done
    projection=dict(
        type="ConvSingleProj",
        in_channels=2048,
        out_channels=256,
        num_convs=1,
        conv_cfg=dict(kernel_size=1, padding=0),
        norm_cfg=dict(type="GN", num_groups=32),
        act_cfg=None,
    ),
    transformer=dict(
        type="TadTRTransformer",
        num_proposals=40,
        num_classes=20,
        with_act_reg=True,
        roi_size=16,
        roi_extend_ratio=0.25,
        aux_loss=True,
        position_embedding=dict(
            type="PositionEmbeddingSine",
            num_pos_feats=256,
            temperature=10000,
            offset=-0.5,
            normalize=True,
        ),
        encoder=dict(
            type="DeformableDETREncoder",
            embed_dim=256,
            num_heads=8,
            num_points=4,
            attn_dropout=0.1,
            ffn_dim=1024,
            ffn_dropout=0.1,
            num_layers=4,
            num_feature_levels=1,
            post_norm=False,
        ),
        decoder=dict(
            type="DeformableDETRDecoder",
            embed_dim=256,
            num_heads=8,
            num_points=4,
            attn_dropout=0.1,
            ffn_dim=1024,
            ffn_dropout=0.1,
            num_layers=4,
            num_feature_levels=1,
            return_intermediate=True,
        ),
        loss=dict(
            type="TadTRSetCriterion",
            num_classes=20,
            matcher=dict(
                type="HungarianMatcher",
                cost_class=6.0,
                cost_bbox=5.0,
                cost_giou=2.0,
                cost_class_type="focal_loss_cost",  # ce_cost, focal_loss_cost
                iou_type="iou",
                use_multi_class=False,
            ),
            loss_class_type="focal_loss",  # ce_loss, focal_loss
            weight_dict=dict(
                loss_class=2.0,
                loss_bbox=5.0,
                loss_iou=2.0,
                loss_actionness=4.0,
            ),
            use_multi_class=False,
        ),
    ),
)