[InferenceClient] Keep generate_kwargs instead of generation_parameters (#2711)

* deprecate instead of removing generate_kwargs

* remove generation_parameters

* remove generation_parameters from other tasks

Author: hanouticelina

src/huggingface_hub/inference/_generated/types/automatic_speech_recognition.py

@@ -76,10 +76,11 @@ class AutomaticSpeechRecognitionGenerationParameters(BaseInferenceType):
 class AutomaticSpeechRecognitionParameters(BaseInferenceType):
     """Additional inference parameters for Automatic Speech Recognition"""
 
-    generation_parameters: Optional[AutomaticSpeechRecognitionGenerationParameters] = None
-    """Parametrization of the text generation process"""
     return_timestamps: Optional[bool] = None
     """Whether to output corresponding timestamps with the generated text"""
+    # Will be deprecated in the future when the renaming to `generation_parameters` is implemented in transformers
+    generate_kwargs: Optional[AutomaticSpeechRecognitionGenerationParameters] = None
+    """Parametrization of the text generation process"""
 
 
 @dataclass

src/huggingface_hub/inference/_generated/types/image_to_text.py

@@ -76,10 +76,11 @@ class ImageToTextGenerationParameters(BaseInferenceType):
 class ImageToTextParameters(BaseInferenceType):
     """Additional inference parameters for Image To Text"""
 
-    generation_parameters: Optional[ImageToTextGenerationParameters] = None
-    """Parametrization of the text generation process"""
     max_new_tokens: Optional[int] = None
     """The amount of maximum tokens to generate."""
+    # Will be deprecated in the future when the renaming to `generation_parameters` is implemented in transformers
+    generate_kwargs: Optional[ImageToTextGenerationParameters] = None
+    """Parametrization of the text generation process"""
 
 
 @dataclass

src/huggingface_hub/inference/_generated/types/text_to_audio.py

@@ -76,7 +76,8 @@ class TextToAudioGenerationParameters(BaseInferenceType):
 class TextToAudioParameters(BaseInferenceType):
     """Additional inference parameters for Text To Audio"""
 
-    generation_parameters: Optional[TextToAudioGenerationParameters] = None
+    # Will be deprecated in the future when the renaming to `generation_parameters` is implemented in transformers
+    generate_kwargs: Optional[TextToAudioGenerationParameters] = None
     """Parametrization of the text generation process"""
 
 

src/huggingface_hub/inference/_generated/types/text_to_speech.py

@@ -76,7 +76,8 @@ class TextToSpeechGenerationParameters(BaseInferenceType):
 class TextToSpeechParameters(BaseInferenceType):
     """Additional inference parameters for Text To Speech"""
 
-    generation_parameters: Optional[TextToSpeechGenerationParameters] = None
+    # Will be deprecated in the future when the renaming to `generation_parameters` is implemented in transformers
+    generate_kwargs: Optional[TextToSpeechGenerationParameters] = None
     """Parametrization of the text generation process"""