Fix some bugs for two stage training of deformable detr

src/transformers/models/deformable_detr/modeling_deformable_detr.py

@@ -1550,7 +1550,7 @@ def get_valid_ratio(self, mask):
     def get_proposal_pos_embed(self, proposals):
         """Get the position embedding of the proposals."""
 
-        num_pos_feats = 128
+        num_pos_feats = self.config.d_model // 2
         temperature = 10000
         scale = 2 * math.pi
 
@@ -1969,12 +1969,11 @@ def forward(
             outputs_coord = outputs_coord_logits.sigmoid()
             outputs_classes.append(outputs_class)
             outputs_coords.append(outputs_coord)
-        # Keep batch_size as first dimension
-        outputs_class = torch.stack(outputs_classes, dim=1)
-        outputs_coord = torch.stack(outputs_coords, dim=1)
+        outputs_class = torch.stack(outputs_classes)
+        outputs_coord = torch.stack(outputs_coords)
 
-        logits = outputs_class[:, -1]
-        pred_boxes = outputs_coord[:, -1]
+        logits = outputs_class[-1]
+        pred_boxes = outputs_coord[-1]
 
         loss, loss_dict, auxiliary_outputs = None, None, None
         if labels is not None:
@@ -2000,7 +1999,7 @@ def forward(
                 outputs_loss["auxiliary_outputs"] = auxiliary_outputs
             if self.config.two_stage:
                 enc_outputs_coord = outputs.enc_outputs_coord_logits.sigmoid()
-                outputs["enc_outputs"] = {"pred_logits": outputs.enc_outputs_class, "pred_boxes": enc_outputs_coord}
+                outputs_loss["enc_outputs"] = {"logits": outputs.enc_outputs_class, "pred_boxes": enc_outputs_coord}
 
             loss_dict = criterion(outputs_loss, labels)
             # Fourth: compute total loss, as a weighted sum of the various losses
@@ -2232,7 +2231,7 @@ def forward(self, outputs, targets):
                 List of dicts, such that `len(targets) == batch_size`. The expected keys in each dict depends on the
                 losses applied, see each loss' doc.
         """
-        outputs_without_aux = {k: v for k, v in outputs.items() if k != "auxiliary_outputs"}
+        outputs_without_aux = {k: v for k, v in outputs.items() if k != "auxiliary_outputs" and k != "enc_outputs"}
 
         # Retrieve the matching between the outputs of the last layer and the targets
         indices = self.matcher(outputs_without_aux, targets)
@@ -2264,14 +2263,10 @@ def forward(self, outputs, targets):
             enc_outputs = outputs["enc_outputs"]
             bin_targets = copy.deepcopy(targets)
             for bt in bin_targets:
-                bt["labels"] = torch.zeros_like(bt["labels"])
+                bt["class_labels"] = torch.zeros_like(bt["class_labels"])
             indices = self.matcher(enc_outputs, bin_targets)
             for loss in self.losses:
-                kwargs = {}
-                if loss == "labels":
-                    # Logging is enabled only for the last layer
-                    kwargs["log"] = False
-                l_dict = self.get_loss(loss, enc_outputs, bin_targets, indices, num_boxes, **kwargs)
+                l_dict = self.get_loss(loss, enc_outputs, bin_targets, indices, num_boxes)
                 l_dict = {k + "_enc": v for k, v in l_dict.items()}
                 losses.update(l_dict)
 

src/transformers/models/deta/modeling_deta.py

@@ -1455,7 +1455,7 @@ def get_valid_ratio(self, mask):
     def get_proposal_pos_embed(self, proposals):
         """Get the position embedding of the proposals."""
 
-        num_pos_feats = 128
+        num_pos_feats = self.config.d_model // 2
         temperature = 10000
         scale = 2 * math.pi
 

tests/models/deformable_detr/test_modeling_deformable_detr.py

@@ -544,6 +544,21 @@ def test_initialization(self):
                         msg=f"Parameter {name} of model {model_class} seems not properly initialized",
                     )
 
+    def test_two_stage_training(self):
+        model_class = DeformableDetrForObjectDetection
+        config, inputs_dict = self.model_tester.prepare_config_and_inputs_for_common()
+        config.return_dict = True
+        config.two_stage = True
+        config.auxiliary_loss = True
+        config.with_box_refine = True
+
+        model = model_class(config)
+        model.to(torch_device)
+        model.train()
+        inputs = self._prepare_for_class(inputs_dict, model_class, return_labels=True)
+        loss = model(**inputs).loss
+        loss.backward()
+
 
 TOLERANCE = 1e-4