[Inference Client] Add task parameters and a maintenance script of th…

…ese parameters (#2561)

* Add additional parameters to Inference Client tasks

* Add and run task params generation script

* Add back missing test

* Add comments to parameters generation script

* Fix shared classes imports + text-to-speech task

* Satisfy end-of-file-fixer hook

* Move helper function to avoid duplicates across scripts

* Rename helper function for more clarity

* Fix bug in node traversing

* Add comments

* improve docstring formatting

* fixes post-review

* Remove aliases from reference package

* Fix small bug and regenerate task parameters

.github/workflows/python-quality.yml

@@ -42,6 +42,7 @@ jobs:
       - run: .venv/bin/python utils/check_static_imports.py
       - run: .venv/bin/python utils/generate_async_inference_client.py
       - run: .venv/bin/python utils/generate_inference_types.py
+      - run: .venv/bin/python utils/generate_task_parameters.py
 
       # Run type checking at least on huggingface_hub root file to check all modules
       # that can be lazy-loaded actually exist.

Makefile

@@ -11,6 +11,7 @@ quality:
 	python utils/check_contrib_list.py
 	python utils/check_static_imports.py
 	python utils/generate_async_inference_client.py
+
 	mypy src
 
 style:
@@ -20,11 +21,14 @@ style:
 	python utils/check_static_imports.py --update
 	python utils/generate_async_inference_client.py --update
 
-inference_types_check:
+inference_check:
 	python utils/generate_inference_types.py
+	python utils/generate_task_parameters.py
 
-inference_types_update:
+inference_update:
 	python utils/generate_inference_types.py --update
+	python utils/generate_task_parameters.py --update
+
 
 repocard:
 	python utils/push_repocard_examples.py

docs/source/en/package_reference/inference_types.md

@@ -398,5 +398,3 @@ This part of the lib is still under development and will be improved in future r
 [[autodoc]] huggingface_hub.ZeroShotObjectDetectionInputData
 
 [[autodoc]] huggingface_hub.ZeroShotObjectDetectionOutputElement
-
-

docs/source/ko/package_reference/inference_types.md

@@ -397,5 +397,3 @@ rendered properly in your Markdown viewer.
 [[autodoc]] huggingface_hub.ZeroShotObjectDetectionInputData
 
 [[autodoc]] huggingface_hub.ZeroShotObjectDetectionOutputElement
-
-

setup.py

@@ -95,6 +95,7 @@ def get_version() -> str:
 extras["quality"] = [
     "ruff>=0.5.0",
     "mypy==1.5.1",
+    "libcst==1.4.0",
 ]
 
 extras["all"] = extras["testing"] + extras["quality"] + extras["typing"]

src/huggingface_hub/__init__.py

@@ -281,9 +281,11 @@
     "inference._generated.types": [
         "AudioClassificationInput",
         "AudioClassificationOutputElement",
+        "AudioClassificationOutputTransform",
         "AudioClassificationParameters",
         "AudioToAudioInput",
         "AudioToAudioOutputElement",
+        "AutomaticSpeechRecognitionEarlyStoppingEnum",
         "AutomaticSpeechRecognitionGenerationParameters",
         "AutomaticSpeechRecognitionInput",
         "AutomaticSpeechRecognitionOutput",
@@ -328,6 +330,7 @@
         "FillMaskParameters",
         "ImageClassificationInput",
         "ImageClassificationOutputElement",
+        "ImageClassificationOutputTransform",
         "ImageClassificationParameters",
         "ImageSegmentationInput",
         "ImageSegmentationOutputElement",
@@ -336,6 +339,7 @@
         "ImageToImageOutput",
         "ImageToImageParameters",
         "ImageToImageTargetSize",
+        "ImageToTextEarlyStoppingEnum",
         "ImageToTextGenerationParameters",
         "ImageToTextInput",
         "ImageToTextOutput",
@@ -361,6 +365,7 @@
         "Text2TextGenerationParameters",
         "TextClassificationInput",
         "TextClassificationOutputElement",
+        "TextClassificationOutputTransform",
         "TextClassificationParameters",
         "TextGenerationInput",
         "TextGenerationInputGenerateParameters",
@@ -373,6 +378,7 @@
         "TextGenerationStreamOutput",
         "TextGenerationStreamOutputStreamDetails",
         "TextGenerationStreamOutputToken",
+        "TextToAudioEarlyStoppingEnum",
         "TextToAudioGenerationParameters",
         "TextToAudioInput",
         "TextToAudioOutput",
@@ -381,6 +387,7 @@
         "TextToImageOutput",
         "TextToImageParameters",
         "TextToImageTargetSize",
+        "TextToSpeechEarlyStoppingEnum",
         "TextToSpeechGenerationParameters",
         "TextToSpeechInput",
         "TextToSpeechOutput",
@@ -394,6 +401,7 @@
         "TranslationParameters",
         "VideoClassificationInput",
         "VideoClassificationOutputElement",
+        "VideoClassificationOutputTransform",
         "VideoClassificationParameters",
         "VisualQuestionAnsweringInput",
         "VisualQuestionAnsweringInputData",
@@ -796,9 +804,11 @@ def __dir__():
     from .inference._generated.types import (
         AudioClassificationInput,  # noqa: F401
         AudioClassificationOutputElement,  # noqa: F401
+        AudioClassificationOutputTransform,  # noqa: F401
         AudioClassificationParameters,  # noqa: F401
         AudioToAudioInput,  # noqa: F401
         AudioToAudioOutputElement,  # noqa: F401
+        AutomaticSpeechRecognitionEarlyStoppingEnum,  # noqa: F401
         AutomaticSpeechRecognitionGenerationParameters,  # noqa: F401
         AutomaticSpeechRecognitionInput,  # noqa: F401
         AutomaticSpeechRecognitionOutput,  # noqa: F401
@@ -843,6 +853,7 @@ def __dir__():
         FillMaskParameters,  # noqa: F401
         ImageClassificationInput,  # noqa: F401
         ImageClassificationOutputElement,  # noqa: F401
+        ImageClassificationOutputTransform,  # noqa: F401
         ImageClassificationParameters,  # noqa: F401
         ImageSegmentationInput,  # noqa: F401
         ImageSegmentationOutputElement,  # noqa: F401
@@ -851,6 +862,7 @@ def __dir__():
         ImageToImageOutput,  # noqa: F401
         ImageToImageParameters,  # noqa: F401
         ImageToImageTargetSize,  # noqa: F401
+        ImageToTextEarlyStoppingEnum,  # noqa: F401
         ImageToTextGenerationParameters,  # noqa: F401
         ImageToTextInput,  # noqa: F401
         ImageToTextOutput,  # noqa: F401
@@ -876,6 +888,7 @@ def __dir__():
         Text2TextGenerationParameters,  # noqa: F401
         TextClassificationInput,  # noqa: F401
         TextClassificationOutputElement,  # noqa: F401
+        TextClassificationOutputTransform,  # noqa: F401
         TextClassificationParameters,  # noqa: F401
         TextGenerationInput,  # noqa: F401
         TextGenerationInputGenerateParameters,  # noqa: F401
@@ -888,6 +901,7 @@ def __dir__():
         TextGenerationStreamOutput,  # noqa: F401
         TextGenerationStreamOutputStreamDetails,  # noqa: F401
         TextGenerationStreamOutputToken,  # noqa: F401
+        TextToAudioEarlyStoppingEnum,  # noqa: F401
         TextToAudioGenerationParameters,  # noqa: F401
         TextToAudioInput,  # noqa: F401
         TextToAudioOutput,  # noqa: F401
@@ -896,6 +910,7 @@ def __dir__():
         TextToImageOutput,  # noqa: F401
         TextToImageParameters,  # noqa: F401
         TextToImageTargetSize,  # noqa: F401
+        TextToSpeechEarlyStoppingEnum,  # noqa: F401
         TextToSpeechGenerationParameters,  # noqa: F401
         TextToSpeechInput,  # noqa: F401
         TextToSpeechOutput,  # noqa: F401
@@ -909,6 +924,7 @@ def __dir__():
         TranslationParameters,  # noqa: F401
         VideoClassificationInput,  # noqa: F401
         VideoClassificationOutputElement,  # noqa: F401
+        VideoClassificationOutputTransform,  # noqa: F401
         VideoClassificationParameters,  # noqa: F401
         VisualQuestionAnsweringInput,  # noqa: F401
         VisualQuestionAnsweringInputData,  # noqa: F401