Generate: legacy mode is only triggered when `generation_config` is untouched (#25962)

docs/source/en/generation_strategies.md

@@ -55,12 +55,10 @@ When you load a model explicitly, you can inspect the generation configuration t
 >>> from transformers import AutoModelForCausalLM
 
 >>> model = AutoModelForCausalLM.from_pretrained("distilgpt2")
->>> model.generation_config  # doctest: +IGNORE_RESULT
+>>> model.generation_config
 GenerationConfig {
-    "_from_model_config": true,
     "bos_token_id": 50256,
     "eos_token_id": 50256,
-    "transformers_version": "4.26.0.dev0"
 }
 ```
 

src/transformers/generation/configuration_utils.py

@@ -34,6 +34,7 @@ from ..utils import (
 
 
 logger = logging.get_logger(__name__)
+METADATA_FIELDS = ("_from_model_config", "_commit_hash", "_original_object_hash", "transformers_version")
 
 
 class GenerationConfig(PushToHubMixin):
@@ -315,20 +316,19 @@ class GenerationConfig(PushToHubMixin):
         # Validate the values of the attributes
         self.validate(is_init=True)
 
+    def __hash__(self):
+        return hash(self.to_json_string(ignore_metadata=True))
+
     def __eq__(self, other):
         if not isinstance(other, GenerationConfig):
             return False
 
-        self_dict = self.__dict__.copy()
-        other_dict = other.__dict__.copy()
-        # ignore metadata
-        for metadata_field in ("_from_model_config", "_commit_hash", "transformers_version"):
-            self_dict.pop(metadata_field, None)
-            other_dict.pop(metadata_field, None)
-        return self_dict == other_dict
+        self_without_metadata = self.to_json_string(use_diff=False, ignore_metadata=True)
+        other_without_metadata = other.to_json_string(use_diff=False, ignore_metadata=True)
+        return self_without_metadata == other_without_metadata
 
     def __repr__(self):
-        return f"{self.__class__.__name__} {self.to_json_string()}"
+        return f"{self.__class__.__name__} {self.to_json_string(ignore_metadata=True)}"
 
     def validate(self, is_init=False):
         """
@@ -729,7 +729,9 @@ class GenerationConfig(PushToHubMixin):
         else:
             logger.info(f"loading configuration file {configuration_file} from cache at {resolved_config_file}")
 
-        return cls.from_dict(config_dict, **kwargs)
+        config = cls.from_dict(config_dict, **kwargs)
+        config._original_object_hash = hash(config)  # Hash to detect whether the instance was modified
+        return config
 
     @classmethod
     def _dict_from_json_file(cls, json_file: Union[str, os.PathLike]):
@@ -814,8 +816,12 @@ class GenerationConfig(PushToHubMixin):
             `Dict[str, Any]`: Dictionary of all the attributes that make up this configuration instance.
         """
         output = copy.deepcopy(self.__dict__)
+
+        # Fields to ignore at serialization time
         if "_commit_hash" in output:
             del output["_commit_hash"]
+        if "_original_object_hash" in output:
+            del output["_original_object_hash"]
 
         # Transformers version when serializing this file
         output["transformers_version"] = __version__
@@ -823,7 +829,7 @@ class GenerationConfig(PushToHubMixin):
         self.dict_torch_dtype_to_str(output)
         return output
 
-    def to_json_string(self, use_diff: bool = True) -> str:
+    def to_json_string(self, use_diff: bool = True, ignore_metadata: bool = False) -> str:
         """
         Serializes this instance to a JSON string.
 
@@ -831,6 +837,8 @@ class GenerationConfig(PushToHubMixin):
             use_diff (`bool`, *optional*, defaults to `True`):
                 If set to `True`, only the difference between the config instance and the default `GenerationConfig()`
                 is serialized to JSON string.
+            ignore_metadata (`bool`, *optional*, defaults to `False`):
+                Whether to ignore the metadata fields present in the instance
 
         Returns:
             `str`: String containing all the attributes that make up this configuration instance in JSON format.
@@ -839,6 +847,11 @@ class GenerationConfig(PushToHubMixin):
             config_dict = self.to_diff_dict()
         else:
             config_dict = self.to_dict()
+
+        if ignore_metadata:
+            for metadata_field in METADATA_FIELDS:
+                config_dict.pop(metadata_field, None)
+
         return json.dumps(config_dict, indent=2, sort_keys=True) + "\n"
 
     def to_json_file(self, json_file_path: Union[str, os.PathLike], use_diff: bool = True):
@@ -882,6 +895,7 @@ class GenerationConfig(PushToHubMixin):
                     if attr in decoder_config and getattr(config, attr) == getattr(default_generation_config, attr):
                         setattr(config, attr, decoder_config[attr])
 
+        config._original_object_hash = hash(config)  # Hash to detect whether the instance was modified
         return config
 
     def update(self, **kwargs):

src/transformers/generation/flax_utils.py

@@ -310,16 +310,20 @@ class FlaxGenerationMixin:
 
         # priority: `generation_config` argument > `model.generation_config` (the default generation config)
         if generation_config is None:
-            # legacy: users may modify the model configuration to control generation -- update the generation config
-            # model attribute accordingly, if it was created from the model config
-            if self.generation_config._from_model_config:
+            # legacy: users may modify the model configuration to control generation. To trigger this legacy behavior,
+            # two conditions must be met
+            # 1) the generation config must have been created from the model config (`_from_model_config` field);
+            # 2) the generation config must have seen no modification since its creation (the hash is the same).
+            if self.generation_config._from_model_config and self.generation_config._original_object_hash == hash(
+                self.generation_config
+            ):
                 new_generation_config = GenerationConfig.from_model_config(self.config)
                 if new_generation_config != self.generation_config:
                     warnings.warn(
                         "You have modified the pretrained model configuration to control generation. This is a"
                         " deprecated strategy to control generation and will be removed soon, in a future version."
-                        " Please use a generation configuration file (see"
-                        " https://huggingface.co/docs/transformers/main_classes/text_generation )"
+                        " Please use and modify the model generation configuration (see"
+                        " https://huggingface.co/docs/transformers/generation_strategies#default-text-generation-configuration )"
                     )
                     self.generation_config = new_generation_config
             generation_config = self.generation_config

src/transformers/generation/tf_utils.py

@@ -716,16 +716,20 @@ class TFGenerationMixin:
 
         # priority: `generation_config` argument > `model.generation_config` (the default generation config)
         if generation_config is None:
-            # legacy: users may modify the model configuration to control generation -- update the generation config
-            # model attribute accordingly, if it was created from the model config
-            if self.generation_config._from_model_config:
+            # legacy: users may modify the model configuration to control generation. To trigger this legacy behavior,
+            # two conditions must be met
+            # 1) the generation config must have been created from the model config (`_from_model_config` field);
+            # 2) the generation config must have seen no modification since its creation (the hash is the same).
+            if self.generation_config._from_model_config and self.generation_config._original_object_hash == hash(
+                self.generation_config
+            ):
                 new_generation_config = GenerationConfig.from_model_config(self.config)
                 if new_generation_config != self.generation_config:
                     warnings.warn(
                         "You have modified the pretrained model configuration to control generation. This is a"
                         " deprecated strategy to control generation and will be removed soon, in a future version."
-                        " Please use a generation configuration file (see"
-                        " https://huggingface.co/docs/transformers/main_classes/text_generation )"
+                        " Please use and modify the model generation configuration (see"
+                        " https://huggingface.co/docs/transformers/generation_strategies#default-text-generation-configuration )"
                     )
                     self.generation_config = new_generation_config
             generation_config = self.generation_config

src/transformers/generation/utils.py

@@ -1409,16 +1409,20 @@ class GenerationMixin:
 
         # priority: `generation_config` argument > `model.generation_config` (the default generation config)
         if generation_config is None:
-            # legacy: users may modify the model configuration to control generation -- update the generation config
-            # model attribute accordingly, if it was created from the model config
-            if self.generation_config._from_model_config:
+            # legacy: users may modify the model configuration to control generation. To trigger this legacy behavior,
+            # two conditions must be met
+            # 1) the generation config must have been created from the model config (`_from_model_config` field);
+            # 2) the generation config must have seen no modification since its creation (the hash is the same).
+            if self.generation_config._from_model_config and self.generation_config._original_object_hash == hash(
+                self.generation_config
+            ):
                 new_generation_config = GenerationConfig.from_model_config(self.config)
                 if new_generation_config != self.generation_config:
                     warnings.warn(
                         "You have modified the pretrained model configuration to control generation. This is a"
                         " deprecated strategy to control generation and will be removed soon, in a future version."
-                        " Please use a generation configuration file (see"
-                        " https://huggingface.co/docs/transformers/main_classes/text_generation )"
+                        " Please use and modify the model generation configuration (see"
+                        " https://huggingface.co/docs/transformers/generation_strategies#default-text-generation-configuration )"
                     )
                     self.generation_config = new_generation_config
             generation_config = self.generation_config

tests/generation/test_utils.py

@@ -2880,7 +2880,7 @@ class GenerationIntegrationTests(unittest.TestCase, GenerationIntegrationTestsMi
 
         # Generation config max_length != 20 -> no warning
         with warnings.catch_warnings(record=True) as warning_list:
+            # generation_config is modified -> legacy mode is disabled = generation_config takes precedence
             model.generation_config.max_length = 10
-            model.generation_config._from_model_config = False  # otherwise model.config.max_length=20 takes precedence
             model.generate(input_ids)
             self.assertEqual(len(warning_list), 0)