[Fix] refine config of RT-DETR-L

paddlex/inference/models/object_detection/modeling/rt_detr.py

@@ -181,77 +181,107 @@ def __call__(self, head_out, im_shape, scale_factor, pad_shape):
 class RTDETRConfig(PretrainedConfig):
     def __init__(
         self,
-        arch,
-        return_idx,
-        freeze_stem_only,
-        freeze_at,
-        freeze_norm,
-        lr_mult_list,
-        hidden_dim,
-        use_encoder_idx,
-        num_encoder_layers,
-        el_d_model,
-        el_nhead,
-        el_dim_feedforward,
-        el_dropout,
-        el_activation,
-        expansion,
-        tf_num_queries,
-        tf_position_embed_type,
-        tf_feat_strides,
-        tf_num_levels,
-        tf_nhead,
-        tf_num_decoder_layers,
-        tf_backbone_feat_channels,
-        tf_dim_feedforward,
-        tf_dropout,
-        tf_activation,
-        tf_num_denoising,
-        tf_label_noise_ratio,
-        tf_box_noise_scale,
-        tf_learnt_init_query,
-        loss_coeff,
-        aux_loss,
-        use_vfl,
-        matcher_coeff,
-        num_top_queries,
-        use_focal_loss,
+        initializer_range=0.01,
+        initializer_bias_prior_prob=None,
+        layer_norm_eps=1e-5,
+        batch_norm_eps=1e-5,
+        # backbone
+        backbone_config=None,
+        freeze_backbone_batch_norms=True,
+        # encoder HybridEncoder
+        encoder_hidden_dim=256,
+        encoder_in_channels=[512, 1024, 2048],
+        feat_strides=[8, 16, 32],
+        encoder_layers=1,
+        encoder_ffn_dim=1024,
+        encoder_attention_heads=8,
+        dropout=0.0,
+        activation_dropout=0.0,
+        encode_proj_layers=[2],
+        positional_encoding_temperature=10000,
+        encoder_activation_function="gelu",
+        activation_function="silu",
+        eval_size=None,
+        normalize_before=False,
+        hidden_expansion=1.0,
+        # decoder RTDetrTransformer
+        d_model=256,
+        num_queries=300,
+        decoder_in_channels=[256, 256, 256],
+        decoder_ffn_dim=1024,
+        num_feature_levels=3,
+        decoder_n_points=4,
+        decoder_layers=6,
+        decoder_attention_heads=8,
+        decoder_activation_function="relu",
+        attention_dropout=0.0,
+        num_denoising=100,
+        label_noise_ratio=0.5,
+        box_noise_scale=1.0,
+        learn_initial_query=False,
+        anchor_image_size=None,
+        disable_custom_kernels=True,
+        with_box_refine=True,
+        is_encoder_decoder=True,
+        # Loss
+        matcher_alpha=0.25,
+        matcher_gamma=2.0,
+        matcher_class_cost=2.0,
+        matcher_bbox_cost=5.0,
+        matcher_giou_cost=2.0,
+        use_focal_loss=True,
+        auxiliary_loss=True,
+        focal_loss_alpha=0.75,
+        focal_loss_gamma=2.0,
+        weight_loss_vfl=1.0,
+        weight_loss_bbox=5.0,
+        weight_loss_giou=2.0,
+        eos_coefficient=1e-4,
         **kwargs,
     ):
-        self.arch = arch
-        self.return_idx = return_idx
-        self.freeze_stem_only = freeze_stem_only
-        self.freeze_at = freeze_at
-        self.freeze_norm = freeze_norm
-        self.lr_mult_list = lr_mult_list
-        self.hidden_dim = hidden_dim
-        self.use_encoder_idx = use_encoder_idx
-        self.num_encoder_layers = num_encoder_layers
+        if backbone_config["model_type"] != "hgnet_v2":
+            raise RuntimeError(
+                f"There is no dynamic graph implementation for backbone {repr(backbone_config["model_type"])}."
+            )
+        self.arch = backbone_config["arch"]
+        self.freeze_stem_only = backbone_config["freeze_stem_only"]
+        self.freeze_at = backbone_config["freeze_at"]
+        self.freeze_norm = backbone_config["freeze_norm"]
+        self.lr_mult_list = backbone_config["lr_mult_list"]
+        self.return_idx = backbone_config["return_idx"]
+        self.hidden_dim = encoder_hidden_dim
+        self.use_encoder_idx = encode_proj_layers
+        self.num_encoder_layers = encoder_layers
         self.el_d_model = d_model
-        self.el_nhead = nhead
-        self.el_dim_feedforward = dim_feedforward
+        self.el_nhead = encoder_attention_heads
+        self.el_dim_feedforward = encoder_ffn_dim
         self.el_dropout = dropout
-        self.el_activation = activation
-        self.expansion = expansion
+        self.el_activation = encoder_activation_function
+        self.expansion = hidden_expansion
         self.tf_num_queries = num_queries
-        self.tf_position_embed_type = position_embed_type
         self.tf_feat_strides = feat_strides
-        self.tf_num_levels = num_levels
-        self.tf_nhead = nhead
-        self.tf_num_decoder_layers = num_decoder_layers
-        self.tf_backbone_feat_channels = backbone_feat_channels
-        self.tf_dim_feedforward = dim_feedforward
-        self.tf_dropout = dropout
-        self.tf_activation = activation
+        self.tf_num_levels = num_feature_levels
+        self.tf_nhead = decoder_attention_heads
+        self.tf_num_decoder_layers = decoder_layers
+        self.tf_backbone_feat_channels = decoder_in_channels
+        self.tf_dim_feedforward = decoder_ffn_dim
+        self.tf_dropout = attention_dropout
+        self.tf_activation = decoder_activation_function
         self.tf_num_denoising = num_denoising
         self.tf_label_noise_ratio = label_noise_ratio
         self.tf_box_noise_scale = box_noise_scale
-        self.tf_learnt_init_query = learnt_init_query
-        self.loss_coeff = loss_coeff
-        self.aux_loss = aux_loss
-        self.use_vfl = use_vfl
-        self.matcher_coeff = matcher_coeff
-        self.num_top_queries = num_top_queries
+        self.tf_learnt_init_query = learn_initial_query
+        self.loss_coeff = {
+            "class": weight_loss_vfl,
+            "bbox": weight_loss_bbox,
+            "giou": weight_loss_giou
+        }
+        self.aux_loss = auxiliary_loss
+        self.matcher_coeff = {
+            "class": matcher_class_cost,
+            "bbox": matcher_bbox_cost,
+            "giou": matcher_giou_cost
+        }
         self.use_focal_loss = use_focal_loss
         self.tensor_parallel_degree = 1
 
@@ -286,7 +316,6 @@ def __init__(self, config: RTDETRConfig):
         )
         self.transformer = RTDETRTransformer(
             num_queries=self.config.tf_num_queries,
-            position_embed_type=self.config.tf_position_embed_type,
             feat_strides=self.config.tf_feat_strides,
             backbone_feat_channels=self.config.tf_backbone_feat_channels,
             num_levels=self.config.tf_num_levels,
@@ -304,14 +333,13 @@ def __init__(self, config: RTDETRConfig):
             loss=DINOLoss(
                 loss_coeff=self.config.loss_coeff,
                 aux_loss=self.config.aux_loss,
-                use_vfl=self.config.use_vfl,
                 matcher=HungarianMatcher(
                     matcher_coeff=self.config.matcher_coeff,
                 ),
             )
         )
         self.post_process = DETRPostProcess(
-            num_top_queries=self.config.num_top_queries,
+            num_top_queries=self.config.tf_num_queries,
             use_focal_loss=self.config.use_focal_loss,
         )