huggingface-hub 0.25.2__py3-none-any.whl → 0.26.0rc0__py3-none-any.whl

huggingface_hub/inference/_generated/types/__init__.py

@@ -6,10 +6,12 @@
 from .audio_classification import (
     AudioClassificationInput,
     AudioClassificationOutputElement,
+    AudioClassificationOutputTransform,
     AudioClassificationParameters,
 )
 from .audio_to_audio import AudioToAudioInput, AudioToAudioOutputElement
 from .automatic_speech_recognition import (
+    AutomaticSpeechRecognitionEarlyStoppingEnum,
     AutomaticSpeechRecognitionGenerationParameters,
     AutomaticSpeechRecognitionInput,
     AutomaticSpeechRecognitionOutput,
@@ -24,8 +26,8 @@ from .chat_completion import (
     ChatCompletionInputGrammarType,
     ChatCompletionInputMessage,
     ChatCompletionInputMessageChunk,
-    ChatCompletionInputTool,
-    ChatCompletionInputToolTypeClass,
+    ChatCompletionInputStreamOptions,
+    ChatCompletionInputToolType,
     ChatCompletionInputURL,
     ChatCompletionOutput,
     ChatCompletionOutputComplete,
@@ -44,6 +46,8 @@ from .chat_completion import (
     ChatCompletionStreamOutputLogprob,
     ChatCompletionStreamOutputLogprobs,
     ChatCompletionStreamOutputTopLogprob,
+    ChatCompletionStreamOutputUsage,
+    ToolElement,
 )
 from .depth_estimation import DepthEstimationInput, DepthEstimationOutput
 from .document_question_answering import (
@@ -57,11 +61,18 @@ from .fill_mask import FillMaskInput, FillMaskOutputElement, FillMaskParameters
 from .image_classification import (
     ImageClassificationInput,
     ImageClassificationOutputElement,
+    ImageClassificationOutputTransform,
     ImageClassificationParameters,
 )
 from .image_segmentation import ImageSegmentationInput, ImageSegmentationOutputElement, ImageSegmentationParameters
 from .image_to_image import ImageToImageInput, ImageToImageOutput, ImageToImageParameters, ImageToImageTargetSize
-from .image_to_text import ImageToTextGenerationParameters, ImageToTextInput, ImageToTextOutput, ImageToTextParameters
+from .image_to_text import (
+    ImageToTextEarlyStoppingEnum,
+    ImageToTextGenerationParameters,
+    ImageToTextInput,
+    ImageToTextOutput,
+    ImageToTextParameters,
+)
 from .object_detection import (
     ObjectDetectionBoundingBox,
     ObjectDetectionInput,
@@ -75,14 +86,19 @@ from .question_answering import (
     QuestionAnsweringParameters,
 )
 from .sentence_similarity import SentenceSimilarityInput, SentenceSimilarityInputData
-from .summarization import SummarizationGenerationParameters, SummarizationInput, SummarizationOutput
+from .summarization import SummarizationInput, SummarizationOutput, SummarizationParameters
 from .table_question_answering import (
     TableQuestionAnsweringInput,
     TableQuestionAnsweringInputData,
     TableQuestionAnsweringOutputElement,
 )
 from .text2text_generation import Text2TextGenerationInput, Text2TextGenerationOutput, Text2TextGenerationParameters
-from .text_classification import TextClassificationInput, TextClassificationOutputElement, TextClassificationParameters
+from .text_classification import (
+    TextClassificationInput,
+    TextClassificationOutputElement,
+    TextClassificationOutputTransform,
+    TextClassificationParameters,
+)
 from .text_generation import (
     TextGenerationInput,
     TextGenerationInputGenerateParameters,
@@ -96,17 +112,31 @@ from .text_generation import (
     TextGenerationStreamOutputStreamDetails,
     TextGenerationStreamOutputToken,
 )
-from .text_to_audio import TextToAudioGenerationParameters, TextToAudioInput, TextToAudioOutput, TextToAudioParameters
+from .text_to_audio import (
+    TextToAudioEarlyStoppingEnum,
+    TextToAudioGenerationParameters,
+    TextToAudioInput,
+    TextToAudioOutput,
+    TextToAudioParameters,
+)
 from .text_to_image import TextToImageInput, TextToImageOutput, TextToImageParameters, TextToImageTargetSize
+from .text_to_speech import (
+    TextToSpeechEarlyStoppingEnum,
+    TextToSpeechGenerationParameters,
+    TextToSpeechInput,
+    TextToSpeechOutput,
+    TextToSpeechParameters,
+)
 from .token_classification import (
     TokenClassificationInput,
     TokenClassificationOutputElement,
     TokenClassificationParameters,
 )
-from .translation import TranslationGenerationParameters, TranslationInput, TranslationOutput
+from .translation import TranslationInput, TranslationOutput, TranslationParameters
 from .video_classification import (
     VideoClassificationInput,
     VideoClassificationOutputElement,
+    VideoClassificationOutputTransform,
     VideoClassificationParameters,
 )
 from .visual_question_answering import (

huggingface_hub/inference/_generated/types/audio_classification.py

@@ -4,12 +4,12 @@
 #   - script: https://github.com/huggingface/huggingface.js/blob/main/packages/tasks/scripts/inference-codegen.ts
 #   - specs:  https://github.com/huggingface/huggingface.js/tree/main/packages/tasks/src/tasks.
 from dataclasses import dataclass
-from typing import Any, Literal, Optional
+from typing import Literal, Optional
 
 from .base import BaseInferenceType
 
 
-ClassificationOutputTransform = Literal["sigmoid", "softmax", "none"]
+AudioClassificationOutputTransform = Literal["sigmoid", "softmax", "none"]
 
 
 @dataclass
@@ -18,7 +18,8 @@ class AudioClassificationParameters(BaseInferenceType):
     Additional inference parameters for Audio Classification
     """
 
-    function_to_apply: Optional["ClassificationOutputTransform"] = None
+    function_to_apply: Optional["AudioClassificationOutputTransform"] = None
+    """The function to apply to the output."""
     top_k: Optional[int] = None
     """When specified, limits the output to the top K most probable classes."""
 
@@ -27,8 +28,10 @@ class AudioClassificationParameters(BaseInferenceType):
 class AudioClassificationInput(BaseInferenceType):
     """Inputs for Audio Classification inference"""
 
-    inputs: Any
-    """The input audio data"""
+    inputs: str
+    """The input audio data as a base64-encoded string. If no `parameters` are provided, you can
+    also provide the audio data as a raw bytes payload.
+    """
     parameters: Optional[AudioClassificationParameters] = None
     """Additional inference parameters"""
 

huggingface_hub/inference/_generated/types/automatic_speech_recognition.py

@@ -4,12 +4,12 @@
 #   - script: https://github.com/huggingface/huggingface.js/blob/main/packages/tasks/scripts/inference-codegen.ts
 #   - specs:  https://github.com/huggingface/huggingface.js/tree/main/packages/tasks/src/tasks.
 from dataclasses import dataclass
-from typing import Any, List, Literal, Optional, Union
+from typing import List, Literal, Optional, Union
 
 from .base import BaseInferenceType
 
 
-EarlyStoppingEnum = Literal["never"]
+AutomaticSpeechRecognitionEarlyStoppingEnum = Literal["never"]
 
 
 @dataclass
@@ -20,7 +20,7 @@ class AutomaticSpeechRecognitionGenerationParameters(BaseInferenceType):
 
     do_sample: Optional[bool] = None
     """Whether to use sampling instead of greedy decoding when generating new tokens."""
-    early_stopping: Optional[Union[bool, "EarlyStoppingEnum"]] = None
+    early_stopping: Optional[Union[bool, "AutomaticSpeechRecognitionEarlyStoppingEnum"]] = None
     """Controls the stopping condition for beam-based methods."""
     epsilon_cutoff: Optional[float] = None
     """If set to float strictly between 0 and 1, only tokens with a conditional probability
@@ -40,11 +40,11 @@ class AutomaticSpeechRecognitionGenerationParameters(BaseInferenceType):
     max_length: Optional[int] = None
     """The maximum length (in tokens) of the generated text, including the input."""
     max_new_tokens: Optional[int] = None
-    """The maximum number of tokens to generate. Takes precedence over maxLength."""
+    """The maximum number of tokens to generate. Takes precedence over max_length."""
     min_length: Optional[int] = None
     """The minimum length (in tokens) of the generated text, including the input."""
     min_new_tokens: Optional[int] = None
-    """The minimum number of tokens to generate. Takes precedence over maxLength."""
+    """The minimum number of tokens to generate. Takes precedence over min_length."""
     num_beam_groups: Optional[int] = None
     """Number of groups to divide num_beams into in order to ensure diversity among different
     groups of beams. See [this paper](https://hf.co/papers/1610.02424) for more details.
@@ -90,8 +90,10 @@ class AutomaticSpeechRecognitionParameters(BaseInferenceType):
 class AutomaticSpeechRecognitionInput(BaseInferenceType):
     """Inputs for Automatic Speech Recognition inference"""
 
-    inputs: Any
-    """The input audio data"""
+    inputs: str
+    """The input audio data as a base64-encoded string. If no `parameters` are provided, you can
+    also provide the audio data as a raw bytes payload.
+    """
     parameters: Optional[AutomaticSpeechRecognitionParameters] = None
     """Additional inference parameters"""
 

huggingface_hub/inference/_generated/types/chat_completion.py

@@ -44,13 +44,23 @@ class ChatCompletionInputGrammarType(BaseInferenceType):
     """
 
 
+@dataclass
+class ChatCompletionInputStreamOptions(BaseInferenceType):
+    include_usage: bool
+    """If set, an additional chunk will be streamed before the data: [DONE] message. The usage
+    field on this chunk shows the token usage statistics for the entire request, and the
+    choices field will always be an empty array. All other chunks will also include a usage
+    field, but with a null value.
+    """
+
+
 @dataclass
 class ChatCompletionInputFunctionName(BaseInferenceType):
     name: str
 
 
 @dataclass
-class ChatCompletionInputToolTypeClass(BaseInferenceType):
+class ChatCompletionInputToolType(BaseInferenceType):
     function: Optional[ChatCompletionInputFunctionName] = None
 
 
@@ -62,7 +72,7 @@ class ChatCompletionInputFunctionDefinition(BaseInferenceType):
 
 
 @dataclass
-class ChatCompletionInputTool(BaseInferenceType):
+class ToolElement(BaseInferenceType):
     function: ChatCompletionInputFunctionDefinition
     type: str
 
@@ -121,16 +131,17 @@ class ChatCompletionInput(BaseInferenceType):
     stop: Optional[List[str]] = None
     """Up to 4 sequences where the API will stop generating further tokens."""
     stream: Optional[bool] = None
+    stream_options: Optional[ChatCompletionInputStreamOptions] = None
     temperature: Optional[float] = None
     """What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the
     output more random, while
     lower values like 0.2 will make it more focused and deterministic.
     We generally recommend altering this or `top_p` but not both.
     """
-    tool_choice: Optional[Union[ChatCompletionInputToolTypeClass, str]] = None
+    tool_choice: Optional[Union[ChatCompletionInputToolType, str]] = None
     tool_prompt: Optional[str] = None
     """A prompt to be appended before the tools"""
-    tools: Optional[List[ChatCompletionInputTool]] = None
+    tools: Optional[List[ToolElement]] = None
     """A list of tools the model may call. Currently, only functions are supported as a tool.
     Use this to provide a list of
     functions the model may generate JSON inputs for.
@@ -265,6 +276,13 @@ class ChatCompletionStreamOutputChoice(BaseInferenceType):
     logprobs: Optional[ChatCompletionStreamOutputLogprobs] = None
 
 
+@dataclass
+class ChatCompletionStreamOutputUsage(BaseInferenceType):
+    completion_tokens: int
+    prompt_tokens: int
+    total_tokens: int
+
+
 @dataclass
 class ChatCompletionStreamOutput(BaseInferenceType):
     """Chat Completion Stream Output.
@@ -278,3 +296,4 @@ class ChatCompletionStreamOutput(BaseInferenceType):
     id: str
     model: str
     system_fingerprint: str
+    usage: Optional[ChatCompletionStreamOutputUsage] = None

huggingface_hub/inference/_generated/types/image_classification.py

@@ -4,12 +4,12 @@
 #   - script: https://github.com/huggingface/huggingface.js/blob/main/packages/tasks/scripts/inference-codegen.ts
 #   - specs:  https://github.com/huggingface/huggingface.js/tree/main/packages/tasks/src/tasks.
 from dataclasses import dataclass
-from typing import Any, Literal, Optional
+from typing import Literal, Optional
 
 from .base import BaseInferenceType
 
 
-ClassificationOutputTransform = Literal["sigmoid", "softmax", "none"]
+ImageClassificationOutputTransform = Literal["sigmoid", "softmax", "none"]
 
 
 @dataclass
@@ -18,7 +18,8 @@ class ImageClassificationParameters(BaseInferenceType):
     Additional inference parameters for Image Classification
     """
 
-    function_to_apply: Optional["ClassificationOutputTransform"] = None
+    function_to_apply: Optional["ImageClassificationOutputTransform"] = None
+    """The function to apply to the output."""
     top_k: Optional[int] = None
     """When specified, limits the output to the top K most probable classes."""
 
@@ -27,8 +28,10 @@ class ImageClassificationParameters(BaseInferenceType):
 class ImageClassificationInput(BaseInferenceType):
     """Inputs for Image Classification inference"""
 
-    inputs: Any
-    """The input image data"""
+    inputs: str
+    """The input image data as a base64-encoded string. If no `parameters` are provided, you can
+    also provide the image data as a raw bytes payload.
+    """
     parameters: Optional[ImageClassificationParameters] = None
     """Additional inference parameters"""
 

huggingface_hub/inference/_generated/types/image_segmentation.py

@@ -4,7 +4,7 @@
 #   - script: https://github.com/huggingface/huggingface.js/blob/main/packages/tasks/scripts/inference-codegen.ts
 #   - specs:  https://github.com/huggingface/huggingface.js/tree/main/packages/tasks/src/tasks.
 from dataclasses import dataclass
-from typing import Any, Literal, Optional
+from typing import Literal, Optional
 
 from .base import BaseInferenceType
 
@@ -32,8 +32,10 @@ class ImageSegmentationParameters(BaseInferenceType):
 class ImageSegmentationInput(BaseInferenceType):
     """Inputs for Image Segmentation inference"""
 
-    inputs: Any
-    """The input image data"""
+    inputs: str
+    """The input image data as a base64-encoded string. If no `parameters` are provided, you can
+    also provide the image data as a raw bytes payload.
+    """
     parameters: Optional[ImageSegmentationParameters] = None
     """Additional inference parameters"""
 
@@ -45,8 +47,8 @@ class ImageSegmentationOutputElement(BaseInferenceType):
     """
 
     label: str
-    """The label of the predicted segment"""
-    mask: Any
-    """The corresponding mask as a black-and-white image"""
+    """The label of the predicted segment."""
+    mask: str
+    """The corresponding mask as a black-and-white image (base64-encoded)."""
     score: Optional[float] = None
-    """The score or confidence degreee the model has"""
+    """The score or confidence degree the model has."""

huggingface_hub/inference/_generated/types/image_to_image.py

@@ -11,7 +11,7 @@ from .base import BaseInferenceType
 
 @dataclass
 class ImageToImageTargetSize(BaseInferenceType):
-    """The size in pixel of the output image"""
+    """The size in pixel of the output image."""
 
     height: int
     width: int
@@ -34,15 +34,17 @@ class ImageToImageParameters(BaseInferenceType):
     a higher quality image at the expense of slower inference.
     """
     target_size: Optional[ImageToImageTargetSize] = None
-    """The size in pixel of the output image"""
+    """The size in pixel of the output image."""
 
 
 @dataclass
 class ImageToImageInput(BaseInferenceType):
     """Inputs for Image To Image inference"""
 
-    inputs: Any
-    """The input image data"""
+    inputs: str
+    """The input image data as a base64-encoded string. If no `parameters` are provided, you can
+    also provide the image data as a raw bytes payload.
+    """
     parameters: Optional[ImageToImageParameters] = None
     """Additional inference parameters"""
 
@@ -52,4 +54,4 @@ class ImageToImageOutput(BaseInferenceType):
     """Outputs of inference for the Image To Image task"""
 
     image: Any
-    """The output image"""
+    """The output image returned as raw bytes in the payload."""

huggingface_hub/inference/_generated/types/image_to_text.py

@@ -9,7 +9,7 @@ from typing import Any, Literal, Optional, Union
 from .base import BaseInferenceType
 
 
-EarlyStoppingEnum = Literal["never"]
+ImageToTextEarlyStoppingEnum = Literal["never"]
 
 
 @dataclass
@@ -20,7 +20,7 @@ class ImageToTextGenerationParameters(BaseInferenceType):
 
     do_sample: Optional[bool] = None
     """Whether to use sampling instead of greedy decoding when generating new tokens."""
-    early_stopping: Optional[Union[bool, "EarlyStoppingEnum"]] = None
+    early_stopping: Optional[Union[bool, "ImageToTextEarlyStoppingEnum"]] = None
     """Controls the stopping condition for beam-based methods."""
     epsilon_cutoff: Optional[float] = None
     """If set to float strictly between 0 and 1, only tokens with a conditional probability
@@ -40,11 +40,11 @@ class ImageToTextGenerationParameters(BaseInferenceType):
     max_length: Optional[int] = None
     """The maximum length (in tokens) of the generated text, including the input."""
     max_new_tokens: Optional[int] = None
-    """The maximum number of tokens to generate. Takes precedence over maxLength."""
+    """The maximum number of tokens to generate. Takes precedence over max_length."""
     min_length: Optional[int] = None
     """The minimum length (in tokens) of the generated text, including the input."""
     min_new_tokens: Optional[int] = None
-    """The minimum number of tokens to generate. Takes precedence over maxLength."""
+    """The minimum number of tokens to generate. Takes precedence over min_length."""
     num_beam_groups: Optional[int] = None
     """Number of groups to divide num_beams into in order to ensure diversity among different
     groups of beams. See [this paper](https://hf.co/papers/1610.02424) for more details.

huggingface_hub/inference/_generated/types/object_detection.py

@@ -4,7 +4,7 @@
 #   - script: https://github.com/huggingface/huggingface.js/blob/main/packages/tasks/scripts/inference-codegen.ts
 #   - specs:  https://github.com/huggingface/huggingface.js/tree/main/packages/tasks/src/tasks.
 from dataclasses import dataclass
-from typing import Any, Optional
+from typing import Optional
 
 from .base import BaseInferenceType
 
@@ -23,8 +23,10 @@ class ObjectDetectionParameters(BaseInferenceType):
 class ObjectDetectionInput(BaseInferenceType):
     """Inputs for Object Detection inference"""
 
-    inputs: Any
-    """The input image data"""
+    inputs: str
+    """The input image data as a base64-encoded string. If no `parameters` are provided, you can
+    also provide the image data as a raw bytes payload.
+    """
     parameters: Optional[ObjectDetectionParameters] = None
     """Additional inference parameters"""
 
@@ -36,9 +38,13 @@ class ObjectDetectionBoundingBox(BaseInferenceType):
     """
 
     xmax: int
+    """The x-coordinate of the bottom-right corner of the bounding box."""
     xmin: int
+    """The x-coordinate of the top-left corner of the bounding box."""
     ymax: int
+    """The y-coordinate of the bottom-right corner of the bounding box."""
     ymin: int
+    """The y-coordinate of the top-left corner of the bounding box."""
 
 
 @dataclass
@@ -50,6 +56,6 @@ class ObjectDetectionOutputElement(BaseInferenceType):
     image.
     """
     label: str
-    """The predicted label for the bounding box"""
+    """The predicted label for the bounding box."""
     score: float
-    """The associated score / probability"""
+    """The associated score / probability."""

huggingface_hub/inference/_generated/types/summarization.py

@@ -9,33 +9,31 @@ from typing import Any, Dict, Literal, Optional
 from .base import BaseInferenceType
 
 
-SummarizationGenerationTruncationStrategy = Literal["do_not_truncate", "longest_first", "only_first", "only_second"]
+SummarizationTruncationStrategy = Literal["do_not_truncate", "longest_first", "only_first", "only_second"]
 
 
 @dataclass
-class SummarizationGenerationParameters(BaseInferenceType):
-    """Additional inference parameters
-    Additional inference parameters for Text2text Generation
+class SummarizationParameters(BaseInferenceType):
+    """Additional inference parameters.
+    Additional inference parameters for summarization.
     """
 
     clean_up_tokenization_spaces: Optional[bool] = None
     """Whether to clean up the potential extra spaces in the text output."""
     generate_parameters: Optional[Dict[str, Any]] = None
-    """Additional parametrization of the text generation algorithm"""
-    truncation: Optional["SummarizationGenerationTruncationStrategy"] = None
-    """The truncation strategy to use"""
+    """Additional parametrization of the text generation algorithm."""
+    truncation: Optional["SummarizationTruncationStrategy"] = None
+    """The truncation strategy to use."""
 
 
 @dataclass
 class SummarizationInput(BaseInferenceType):
-    """Inputs for Summarization inference
-    Inputs for Text2text Generation inference
-    """
+    """Inputs for Summarization inference"""
 
     inputs: str
-    """The input text data"""
-    parameters: Optional[SummarizationGenerationParameters] = None
-    """Additional inference parameters"""
+    """The input text to summarize."""
+    parameters: Optional[SummarizationParameters] = None
+    """Additional inference parameters."""
 
 
 @dataclass

huggingface_hub/inference/_generated/types/text_classification.py

@@ -9,18 +9,23 @@ from typing import Literal, Optional
 from .base import BaseInferenceType
 
 
-ClassificationOutputTransform = Literal["sigmoid", "softmax", "none"]
+TextClassificationOutputTransform = Literal["sigmoid", "softmax", "none"]
 
 
 @dataclass
 class TextClassificationParameters(BaseInferenceType):
-    """Additional inference parameters
-    Additional inference parameters for Text Classification
+    """
+    Additional inference parameters for Text Classification.
     """
 
-    function_to_apply: Optional["ClassificationOutputTransform"] = None
+    function_to_apply: Optional["TextClassificationOutputTransform"] = None
+    """
+    The function to apply to the output.
+    """
     top_k: Optional[int] = None
-    """When specified, limits the output to the top K most probable classes."""
+    """
+    When specified, limits the output to the top K most probable classes.
+    """
 
 
 @dataclass

huggingface_hub/inference/_generated/types/text_generation.py

@@ -142,6 +142,7 @@ class TextGenerationOutput(BaseInferenceType):
 class TextGenerationStreamOutputStreamDetails(BaseInferenceType):
     finish_reason: "TextGenerationOutputFinishReason"
     generated_tokens: int
+    input_length: int
     seed: Optional[int] = None
 
 

huggingface_hub/inference/_generated/types/text_to_audio.py

@@ -9,7 +9,7 @@ from typing import Any, Literal, Optional, Union
 from .base import BaseInferenceType
 
 
-EarlyStoppingEnum = Literal["never"]
+TextToAudioEarlyStoppingEnum = Literal["never"]
 
 
 @dataclass
@@ -20,7 +20,7 @@ class TextToAudioGenerationParameters(BaseInferenceType):
 
     do_sample: Optional[bool] = None
     """Whether to use sampling instead of greedy decoding when generating new tokens."""
-    early_stopping: Optional[Union[bool, "EarlyStoppingEnum"]] = None
+    early_stopping: Optional[Union[bool, "TextToAudioEarlyStoppingEnum"]] = None
     """Controls the stopping condition for beam-based methods."""
     epsilon_cutoff: Optional[float] = None
     """If set to float strictly between 0 and 1, only tokens with a conditional probability

huggingface_hub/inference/_generated/types/text_to_image.py

@@ -24,17 +24,19 @@ class TextToImageParameters(BaseInferenceType):
     """
 
     guidance_scale: Optional[float] = None
-    """For diffusion models. A higher guidance scale value encourages the model to generate
-    images closely linked to the text prompt at the expense of lower image quality.
+    """A higher guidance scale value encourages the model to generate images closely linked to
+    the text prompt, but values too high may cause saturation and other artifacts.
     """
     negative_prompt: Optional[List[str]] = None
     """One or several prompt to guide what NOT to include in image generation."""
     num_inference_steps: Optional[int] = None
-    """For diffusion models. The number of denoising steps. More denoising steps usually lead to
-    a higher quality image at the expense of slower inference.
+    """The number of denoising steps. More denoising steps usually lead to a higher quality
+    image at the expense of slower inference.
     """
     scheduler: Optional[str] = None
-    """For diffusion models. Override the scheduler with a compatible one"""
+    """Override the scheduler with a compatible one."""
+    seed: Optional[int] = None
+    """Seed for the random number generator."""
     target_size: Optional[TextToImageTargetSize] = None
     """The size in pixel of the output image"""
 
@@ -44,7 +46,7 @@ class TextToImageInput(BaseInferenceType):
     """Inputs for Text To Image inference"""
 
     inputs: str
-    """The input text data (sometimes called "prompt\""""
+    """The input text data (sometimes called "prompt")"""
     parameters: Optional[TextToImageParameters] = None
     """Additional inference parameters"""
 
@@ -54,4 +56,4 @@ class TextToImageOutput(BaseInferenceType):
     """Outputs of inference for the Text To Image task"""
 
     image: Any
-    """The generated image"""
+    """The generated image returned as raw bytes in the payload."""