additional changes

src/transformers/__init__.py

@@ -5624,9 +5624,7 @@
         RTDetrConfig,
         RTDetrResNetConfig,
     )
-    from .models.rt_detr_v2 import (
-        RtDetrV2Config
-    )
+    from .models.rt_detr_v2 import RtDetrV2Config
     from .models.rwkv import RwkvConfig
     from .models.sam import (
         SamConfig,
@@ -7809,11 +7807,7 @@
             RTDetrResNetBackbone,
             RTDetrResNetPreTrainedModel,
         )
-        from .models.rt_detr_v2 import (
-            RtDetrV2ForObjectDetection,
-            RtDetrV2Model,
-            RtDetrV2PreTrainedModel
-        )
+        from .models.rt_detr_v2 import RtDetrV2ForObjectDetection, RtDetrV2Model, RtDetrV2PreTrainedModel
         from .models.rwkv import (
             RwkvForCausalLM,
             RwkvModel,

src/transformers/models/rt_detr_v2/convert_rt_detr_v2_weights_to_hf.py

@@ -24,7 +24,7 @@
 from PIL import Image
 from torchvision import transforms
 
-from transformers import RTDetrConfig, RTDetrForObjectDetection, RTDetrImageProcessor
+from transformers import RTDetrImageProcessor
 from transformers.models.rt_detr_v2.modular_rt_detr_v2 import RtDetrV2Config, RtDetrV2ForObjectDetection
 from transformers.utils import logging
 
@@ -578,7 +578,6 @@ def convert_rt_detr_checkpoint(model_name, pytorch_dump_folder_path, push_to_hub
 
     # finally, create HuggingFace model and load state dict
     model = RtDetrV2ForObjectDetection(config)
-    breakpoint()
     model.load_state_dict(state_dict)
     model.eval()
 
@@ -610,7 +609,6 @@ def convert_rt_detr_checkpoint(model_name, pytorch_dump_folder_path, push_to_hub
     # Pass image by the model
     outputs = model(pixel_values)
 
-    breakpoint()
     if model_name == "rtdetr_v2_r18vd":
         expected_slice_logits = torch.tensor(
             [[-3.7045, -5.1913, -6.1787], [-4.0106, -9.3450, -5.2043], [-4.1287, -4.7463, -5.8634]]

src/transformers/models/rt_detr_v2/modular_rt_detr_v2.py

@@ -1,39 +1,32 @@
-from ..rt_detr.modeling_rt_detr import (
-    RTDetrDecoderLayer, RTDetrModelOutput, RTDetrObjectDetectionOutput, RTDetrHybridEncoder,
-    RTDetrEncoderLayer, RTDetrConvEncoder, RTDetrMLPPredictionHead, RTDetrDecoderOutput, 
-    RTDetrMultiscaleDeformableAttention, MultiScaleDeformableAttentionFunction, get_contrastive_denoising_training_group,
-    inverse_sigmoid, RTDetrDecoder, RTDetrModel, RTDetrForObjectDetection
-)
-from ...modeling_utils import PreTrainedModel
-from ..rt_detr.configuration_rt_detr import RTDetrConfig
-from ...activations import ACT2FN
 import math
-import os
-import warnings
-from dataclasses import dataclass
-from functools import lru_cache, partial, wraps
-from pathlib import Path
-from typing import Dict, List, Optional, Tuple, Union
+from typing import List, Optional
 
 import torch
 import torch.nn.functional as F
 from torch import Tensor, nn
-from torch.autograd import Function
-from torch.autograd.function import once_differentiable
 
-from ...utils import is_torch_cuda_available
-import torch.nn as nn
+from ...modeling_utils import PreTrainedModel
+from ..rt_detr.configuration_rt_detr import RTDetrConfig
+from ..rt_detr.modeling_rt_detr import (
+    MultiScaleDeformableAttentionFunction,
+    RTDetrDecoder,
+    RTDetrDecoderLayer,
+    RTDetrForObjectDetection,
+    RTDetrModel,
+    RTDetrMultiscaleDeformableAttention,
+)
+
 
 class RtDetrV2Config(RTDetrConfig):
     model_type = "rt_detr_v2"
 
-    def __init__(self, 
+    def __init__(self,
                  decoder_n_levels=3,  # default value
                  decoder_offset_scale=0.5,  # default value
                  **super_kwargs):
         # init the base RTDetrConfig class with additional parameters
         super().__init__(**super_kwargs)
-        
+
         # add the new attributes with the given values or defaults
         self.decoder_n_levels = decoder_n_levels
         self.decoder_offset_scale = decoder_offset_scale
@@ -129,7 +122,7 @@ def __init__(self, config: RtDetrV2Config):
             num_heads=config.decoder_attention_heads,
             n_points=config.decoder_n_points
         )
-        
+
         # V2-specific attributes
         self.offset_scale = config.decoder_offset_scale
         # Initialize n_points list and scale
@@ -152,10 +145,10 @@ def _reset_parameters(self):
         )
         for i in range(self.n_points):
             grid_init[:, :, i, :] *= i + 1
-            
+
         with torch.no_grad():
             self.sampling_offsets.bias = nn.Parameter(grid_init.view(-1))
-            
+
         nn.init.constant_(self.attention_weights.weight.data, 0.0)
         nn.init.constant_(self.attention_weights.bias.data, 0.0)
         nn.init.xavier_uniform_(self.value_proj.weight.data)
@@ -191,12 +184,12 @@ def forward(
         if attention_mask is not None:
             value = value.masked_fill(~attention_mask[..., None], float(0))
         value = value.view(batch_size, sequence_length, self.n_heads, self.d_model // self.n_heads)
-        
+
         # V2-specific sampling offsets shape
         sampling_offsets = self.sampling_offsets(hidden_states).view(
             batch_size, num_queries, self.n_heads, self.n_levels * self.n_points, 2
         )
-        
+
         attention_weights = self.attention_weights(hidden_states).view(
             batch_size, num_queries, self.n_heads, self.n_levels * self.n_points
         )
@@ -253,7 +246,6 @@ def __init__(self, config: RtDetrV2Config):
         self.layers = nn.ModuleList([RtDetrV2DecoderLayer(config) for _ in range(config.decoder_layers)])
 
 
-# could make better use of inheritence
 class RtDetrV2Model(RTDetrModel):
     def __init__(self, config: RtDetrV2Config):
         super().__init__(config)
@@ -277,7 +269,7 @@ class RtDetrV2PreTrainedModel(PreTrainedModel):
     def __init__(self, RTDetrV2Config):
         super().__init__(config)
         self.model = RTDetrV2Model(config)
-    
+
     def _init_weights(self, module):
         # this could be simplified like calling the base class function first then adding new stuff
         """Initalize the weights"""
@@ -292,10 +284,10 @@ def _init_weights(self, module):
         )
         for i in range(self.n_points):
             grid_init[:, :, i, :] *= i + 1
-            
+
         with torch.no_grad():
             self.sampling_offsets.bias = nn.Parameter(grid_init.view(-1))
-            
+
         nn.init.constant_(self.attention_weights.weight.data, 0.0)
         nn.init.constant_(self.attention_weights.bias.data, 0.0)
         nn.init.xavier_uniform_(self.value_proj.weight.data)

tests/models/rtdetrv2/test_modeling_rt_detr_v2.py

@@ -22,8 +22,8 @@
 from parameterized import parameterized
 
 from transformers import (
-    RtDetrV2Config,
     RTDetrImageProcessor,
+    RtDetrV2Config,
     is_torch_available,
     is_vision_available,
 )