Deal with nested configs better in base class (#25237)

* Deal better with nested configs

* Fixes

* More fixes

* Fix last test

* Clean up existing configs

* Remove hack in MPT Config

* Update src/transformers/configuration_utils.py
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

* Fix setting a nested config via dict in the kwargs

* Adapt common test

* Add test for nested config load with dict

---------
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>

src/transformers/models/instructblip/configuration_instructblip.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """ InstructBLIP model configuration"""
 
-import copy
 import os
 from typing import Union
 
@@ -305,7 +304,6 @@ class InstructBlipConfig(PretrainedConfig):
     ```"""
 
     model_type = "instructblip"
-    is_composition = True
 
     def __init__(self, vision_config=None, qformer_config=None, text_config=None, num_query_tokens=32, **kwargs):
         super().__init__(**kwargs)
@@ -358,17 +356,3 @@ class InstructBlipConfig(PretrainedConfig):
             text_config=text_config.to_dict(),
             **kwargs,
         )
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["vision_config"] = self.vision_config.to_dict()
-        output["qformer_config"] = self.qformer_config.to_dict()
-        output["text_config"] = self.text_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/jukebox/configuration_jukebox.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """ Jukebox configuration"""
 
-import copy
 import os
 from typing import List, Union
 
@@ -369,18 +368,6 @@ class JukeboxPriorConfig(PretrainedConfig):
 
         return cls.from_dict(config_dict, **kwargs)
 
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["encoder_config"] = self.encoder_config.to_dict() if self.encoder_config is not None else None
-        output["model_type"] = self.__class__.model_type
-        return output
-
 
 class JukeboxVQVAEConfig(PretrainedConfig):
     """
@@ -561,7 +548,6 @@ class JukeboxConfig(PretrainedConfig):
     """
 
     model_type = "jukebox"
-    is_composition = True
 
     def __init__(
         self,
@@ -620,18 +606,3 @@ class JukeboxConfig(PretrainedConfig):
         """
         prior_config_list = [config.to_dict() for config in prior_configs]
         return cls(prior_config_list=prior_config_list, vqvae_config_dict=vqvae_config.to_dict(), **kwargs)
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        for i, config in enumerate(output.pop("prior_configs")):
-            output[f"prior_{i}"] = config.to_dict()
-
-        output["vqvae_config"] = self.vqvae_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/mask2former/configuration_mask2former.py

@@ -13,7 +13,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """ Mask2Former model configuration"""
-import copy
 from typing import Dict, List, Optional
 
 from ...configuration_utils import PretrainedConfig
@@ -230,15 +229,3 @@ class Mask2FormerConfig(PretrainedConfig):
             backbone_config=backbone_config,
             **kwargs,
         )
-
-    def to_dict(self) -> Dict[str, any]:
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["backbone_config"] = self.backbone_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/maskformer/configuration_maskformer.py

@@ -13,7 +13,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """ MaskFormer model configuration"""
-import copy
 from typing import Dict, Optional
 
 from ...configuration_utils import PretrainedConfig
@@ -200,16 +199,3 @@ class MaskFormerConfig(PretrainedConfig):
             decoder_config=decoder_config,
             **kwargs,
         )
-
-    def to_dict(self) -> Dict[str, any]:
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["backbone_config"] = self.backbone_config.to_dict()
-        output["decoder_config"] = self.decoder_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/mpt/configuration_mpt.py

@@ -13,7 +13,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """ Mpt configuration"""
-import copy
 from typing import TYPE_CHECKING, Optional, Union
 
 
@@ -197,7 +196,6 @@ class MptConfig(PretrainedConfig):
         "hidden_size": "d_model",
         "num_hidden_layers": "n_layers",
     }
-    is_composition = True
 
     def __init__(
         self,
@@ -222,6 +220,11 @@ class MptConfig(PretrainedConfig):
         initializer_range=0.02,
         **kwargs,
     ):
+        if attn_config is None:
+            self.attn_config = MptAttentionConfig()
+        elif isinstance(attn_config, dict):
+            self.attn_config = MptAttentionConfig(**attn_config)
+        else:
             self.attn_config = attn_config
         self.d_model = d_model
         self.n_heads = n_heads
@@ -242,35 +245,3 @@ class MptConfig(PretrainedConfig):
         self.use_cache = use_cache
         self.initializer_range = initializer_range
         super().__init__(**kwargs)
-
-    @property
-    def attn_config(self):
-        return self._attn_config
-
-    @attn_config.setter
-    def attn_config(self, attn_config):
-        if attn_config is None:
-            self._attn_config = MptAttentionConfig()
-        elif isinstance(attn_config, dict):
-            self._attn_config = MptAttentionConfig(**attn_config)
-        elif isinstance(attn_config, MptAttentionConfig):
-            self._attn_config = attn_config
-        else:
-            raise ValueError(
-                f"`attn_config` has to be either a `MptAttentionConfig` or a dictionary. Received: {type(attn_config)}"
-            )
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["attn_config"] = (
-            self._attn_config.to_dict() if not isinstance(self.attn_config, dict) else self.attn_config
-        )
-        del output["_attn_config"]
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/musicgen/configuration_musicgen.py

@@ -13,7 +13,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """ MusicGen model configuration"""
-import copy
 
 from ...configuration_utils import PretrainedConfig
 from ...utils import logging
@@ -227,17 +226,3 @@ class MusicgenConfig(PretrainedConfig):
             decoder=decoder_config.to_dict(),
             **kwargs,
         )
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["text_encoder"] = self.text_encoder.to_dict()
-        output["audio_encoder"] = self.audio_encoder.to_dict()
-        output["decoder"] = self.decoder.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/oneformer/configuration_oneformer.py

@@ -13,7 +13,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """OneFormer model configuration"""
-import copy
 from typing import Dict, Optional
 
 from ...configuration_utils import PretrainedConfig
@@ -250,13 +249,3 @@ class OneFormerConfig(PretrainedConfig):
         self.num_hidden_layers = decoder_layers
 
         super().__init__(**kwargs)
-
-    def to_dict(self) -> Dict[str, any]:
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`]. Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["backbone_config"] = self.backbone_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/owlvit/configuration_owlvit.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """ OWL-ViT model configuration"""
 
-import copy
 import os
 from collections import OrderedDict
 from typing import TYPE_CHECKING, Any, Dict, Mapping, Optional, Union
@@ -274,7 +273,6 @@ class OwlViTConfig(PretrainedConfig):
     """
 
     model_type = "owlvit"
-    is_composition = True
 
     def __init__(
         self,
@@ -332,19 +330,6 @@ class OwlViTConfig(PretrainedConfig):
 
         return cls.from_dict(config_dict, **kwargs)
 
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["text_config"] = self.text_config.to_dict()
-        output["vision_config"] = self.vision_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output
-
 
 class OwlViTOnnxConfig(OnnxConfig):
     @property

src/transformers/models/pix2struct/configuration_pix2struct.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """ Pix2Struct model configuration"""
 
-import copy
 import os
 from typing import Union
 
@@ -338,7 +337,6 @@ class Pix2StructConfig(PretrainedConfig):
     ```"""
 
     model_type = "pix2struct"
-    is_composition = True
 
     def __init__(
         self,
@@ -389,16 +387,3 @@ class Pix2StructConfig(PretrainedConfig):
         """
 
         return cls(text_config=text_config.to_dict(), vision_config=vision_config.to_dict(), **kwargs)
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["text_config"] = self.text_config.to_dict()
-        output["vision_config"] = self.vision_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/rag/configuration_rag.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """ RAG model configuration"""
 
-import copy
 
 from ...configuration_utils import PretrainedConfig
 from ...utils import add_start_docstrings
@@ -179,16 +178,3 @@ class RagConfig(PretrainedConfig):
             [`EncoderDecoderConfig`]: An instance of a configuration object
         """
         return cls(question_encoder=question_encoder_config.to_dict(), generator=generator_config.to_dict(), **kwargs)
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["question_encoder"] = self.question_encoder.to_dict()
-        output["generator"] = self.generator.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/sam/configuration_sam.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """ SAM model configuration"""
 
-import copy
 
 from ...configuration_utils import PretrainedConfig
 from ...utils import logging
@@ -286,7 +285,6 @@ class SamConfig(PretrainedConfig):
     ```"""
 
     model_type = "sam"
-    is_composition = True
 
     def __init__(
         self,
@@ -312,17 +310,3 @@ class SamConfig(PretrainedConfig):
         self.prompt_encoder_config = SamPromptEncoderConfig(**prompt_encoder_config)
         self.mask_decoder_config = SamMaskDecoderConfig(**mask_decoder_config)
         self.initializer_range = initializer_range
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["vision_config"] = self.vision_config.to_dict()
-        output["prompt_encoder_config"] = self.prompt_encoder_config.to_dict()
-        output["mask_decoder_config"] = self.mask_decoder_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/speech_encoder_decoder/configuration_speech_encoder_decoder.py

@@ -14,7 +14,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import copy
 
 from ...configuration_utils import PretrainedConfig
 from ...utils import logging
@@ -106,16 +105,3 @@ class SpeechEncoderDecoderConfig(PretrainedConfig):
         decoder_config.add_cross_attention = True
 
         return cls(encoder=encoder_config.to_dict(), decoder=decoder_config.to_dict(), **kwargs)
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default *to_dict()* from *PretrainedConfig*.
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["encoder"] = self.encoder.to_dict()
-        output["decoder"] = self.decoder.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/table_transformer/configuration_table_transformer.py

@@ -13,9 +13,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """ Table Transformer model configuration"""
-import copy
 from collections import OrderedDict
-from typing import Dict, Mapping
+from typing import Mapping
 
 from packaging import version
 
@@ -237,17 +236,6 @@ class TableTransformerConfig(PretrainedConfig):
     def hidden_size(self) -> int:
         return self.d_model
 
-    def to_dict(self) -> Dict[str, any]:
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`]. Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        if output["backbone_config"] is not None:
-            output["backbone_config"] = self.backbone_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output
-
 
 # Copied from transformers.models.detr.configuration_detr.DetrOnnxConfig
 class TableTransformerOnnxConfig(OnnxConfig):

src/transformers/models/upernet/configuration_upernet.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """ UperNet model configuration"""
 
-import copy
 
 from ...configuration_utils import PretrainedConfig
 from ...utils import logging
@@ -108,13 +107,3 @@ class UperNetConfig(PretrainedConfig):
         self.auxiliary_num_convs = auxiliary_num_convs
         self.auxiliary_concat_input = auxiliary_concat_input
         self.loss_ignore_index = loss_ignore_index
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`]. Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["backbone_config"] = self.backbone_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/vision_encoder_decoder/configuration_vision_encoder_decoder.py

@@ -14,7 +14,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import copy
 from typing import TYPE_CHECKING, Any, Mapping, Optional, OrderedDict
 
 from packaging import version
@@ -114,19 +113,6 @@ class VisionEncoderDecoderConfig(PretrainedConfig):
 
         return cls(encoder=encoder_config.to_dict(), decoder=decoder_config.to_dict(), **kwargs)
 
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default *to_dict()* from *PretrainedConfig*.
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["encoder"] = self.encoder.to_dict()
-        output["decoder"] = self.decoder.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output
-
 
 class VisionEncoderDecoderEncoderOnnxConfig(OnnxConfig):
     torch_onnx_minimum_version = version.parse("1.11")

src/transformers/models/vision_text_dual_encoder/configuration_vision_text_dual_encoder.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """ VisionTextDualEncoder model configuration"""
 
-import copy
 
 from ...configuration_utils import PretrainedConfig
 from ...utils import logging
@@ -113,16 +112,3 @@ class VisionTextDualEncoderConfig(PretrainedConfig):
         """
 
         return cls(vision_config=vision_config.to_dict(), text_config=text_config.to_dict(), **kwargs)
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["vision_config"] = self.vision_config.to_dict()
-        output["text_config"] = self.text_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/vit_hybrid/configuration_vit_hybrid.py

@@ -14,8 +14,6 @@
 # limitations under the License.
 """ ViT Hybrid model configuration"""
 
-import copy
-from typing import Dict
 
 from ...configuration_utils import PretrainedConfig
 from ...utils import logging
@@ -146,13 +144,3 @@ class ViTHybridConfig(PretrainedConfig):
         self.patch_size = patch_size
         self.num_channels = num_channels
         self.qkv_bias = qkv_bias
-
-    def to_dict(self) -> Dict[str, any]:
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`]. Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["backbone_config"] = self.backbone_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

src/transformers/models/x_clip/configuration_x_clip.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """ X-CLIP model configuration"""
 
-import copy
 import os
 from typing import Union
 
@@ -299,7 +298,6 @@ class XCLIPConfig(PretrainedConfig):
     """
 
     model_type = "xclip"
-    is_composition = True
 
     def __init__(
         self,
@@ -417,16 +415,3 @@ class XCLIPConfig(PretrainedConfig):
         """
 
         return cls(text_config=text_config.to_dict(), vision_config=vision_config.to_dict(), **kwargs)
-
-    def to_dict(self):
-        """
-        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
-
-        Returns:
-            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
-        """
-        output = copy.deepcopy(self.__dict__)
-        output["text_config"] = self.text_config.to_dict()
-        output["vision_config"] = self.vision_config.to_dict()
-        output["model_type"] = self.__class__.model_type
-        return output

tests/test_configuration_common.py

@@ -118,7 +118,9 @@ class ConfigTester(object):
 
     def check_config_can_be_init_without_params(self):
         if self.config_class.is_composition:
-            return
+            with self.parent.assertRaises(ValueError):
+                config = self.config_class()
+        else:
             config = self.config_class()
             self.parent.assertIsNotNone(config)
 

tests/test_configuration_utils.py

@@ -210,6 +210,13 @@ class ConfigTestUtils(unittest.TestCase):
                 f" {', '.join(keys_with_defaults)}."
             )
 
+    def test_nested_config_load_from_dict(self):
+        config = AutoConfig.from_pretrained(
+            "hf-internal-testing/tiny-random-CLIPModel", text_config={"num_hidden_layers": 2}
+        )
+        self.assertNotIsInstance(config.text_config, dict)
+        self.assertEqual(config.text_config.__class__.__name__, "CLIPTextConfig")
+
     def test_from_pretrained_subfolder(self):
         with self.assertRaises(OSError):
             # config is in subfolder, the following should not work without specifying the subfolder