@huggingface/tasks 0.13.16 → 0.14.0

package/src/tasks/text-to-audio/spec/output.json

@@ -6,12 +6,13 @@
 	"type": "object",
 	"properties": {
 		"audio": {
-			"description": "The generated audio waveform."
+			"description": "The generated audio waveform.",
+			"comment": "type=binary"
 		},
 		"sampling_rate": {
 			"type": "number",
 			"description": "The sampling rate of the generated audio waveform."
 		}
 	},
-	"required": ["audio", "samplingRate"]
+	"required": ["audio", "sampling_rate"]
 }

package/src/tasks/text-to-image/inference.ts

@@ -3,7 +3,6 @@
  *
  * Using src/scripts/inference-codegen
  */
-
 /**
  * Inputs for Text To Image inference
  */
@@ -18,7 +17,6 @@ export interface TextToImageInput {
 	parameters?: TextToImageParameters;
 	[property: string]: unknown;
 }
-
 /**
  * Additional inference parameters for Text To Image
  */
@@ -29,9 +27,9 @@ export interface TextToImageParameters {
 	 */
 	guidance_scale?: number;
 	/**
-	 * One or several prompt to guide what NOT to include in image generation.
+	 * One prompt to guide what NOT to include in image generation.
 	 */
-	negative_prompt?: string[];
+	negative_prompt?: string;
 	/**
 	 * The number of denoising steps. More denoising steps usually lead to a higher quality
 	 * image at the expense of slower inference.
@@ -51,7 +49,6 @@ export interface TextToImageParameters {
 	target_size?: TargetSize;
 	[property: string]: unknown;
 }
-
 /**
  * The size in pixel of the output image
  */
@@ -60,7 +57,6 @@ export interface TargetSize {
 	width: number;
 	[property: string]: unknown;
 }
-
 /**
  * Outputs of inference for the Text To Image task
  */

package/src/tasks/text-to-image/spec/input.json

@@ -24,11 +24,8 @@
 					"description": "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": {
-					"type": "array",
-					"items": {
-						"type": "string"
-					},
-					"description": "One or several prompt to guide what NOT to include in image generation."
+					"type": "string",
+					"description": "One prompt to guide what NOT to include in image generation."
 				},
 				"num_inference_steps": {
 					"type": "integer",

package/src/tasks/text-to-speech/inference.ts

@@ -1,9 +1,22 @@
+/**
+ * Outputs of inference for the Text To Speech task
+ */
+export interface TextToSpeechOutput {
+	/**
+	 * The generated audio
+	 */
+	audio: Blob;
+	/**
+	 * The sampling rate of the generated audio waveform.
+	 */
+	sampling_rate?: number;
+	[property: string]: unknown;
+}
 /**
  * Inference code generated from the JSON schema spec in ./spec
  *
  * Using src/scripts/inference-codegen
  */
-
 /**
  * Inputs for Text To Speech inference
  */
@@ -18,7 +31,6 @@ export interface TextToSpeechInput {
 	parameters?: TextToSpeechParameters;
 	[property: string]: unknown;
 }
-
 /**
  * Additional inference parameters for Text To Speech
  */
@@ -29,7 +41,6 @@ export interface TextToSpeechParameters {
 	generation_parameters?: GenerationParameters;
 	[property: string]: unknown;
 }
-
 /**
  * Parametrization of the text generation process
  */
@@ -116,26 +127,7 @@ export interface GenerationParameters {
 	use_cache?: boolean;
 	[property: string]: unknown;
 }
-
 /**
  * Controls the stopping condition for beam-based methods.
  */
 export type EarlyStoppingUnion = boolean | "never";
-
-/**
- * Outputs for Text to Speech inference
- *
- * Outputs of inference for the Text To Audio task
- */
-export interface TextToSpeechOutput {
-	/**
-	 * The generated audio waveform.
-	 */
-	audio: unknown;
-	samplingRate: unknown;
-	/**
-	 * The sampling rate of the generated audio waveform.
-	 */
-	sampling_rate?: number;
-	[property: string]: unknown;
-}

package/src/tasks/text-to-speech/spec/output.json

@@ -1,7 +1,18 @@
 {
-	"$ref": "/inference/schemas/text-to-audio/output.json",
 	"$id": "/inference/schemas/text-to-speech/output.json",
 	"$schema": "http://json-schema.org/draft-06/schema#",
+	"description": "Outputs of inference for the Text To Speech task",
 	"title": "TextToSpeechOutput",
-	"description": "Outputs for Text to Speech inference"
+	"type": "object",
+	"properties": {
+		"audio": {
+			"description": "The generated audio",
+			"comment": "type=binary"
+		},
+		"sampling_rate": {
+			"type": "number",
+			"description": "The sampling rate of the generated audio waveform."
+		}
+	},
+	"required": ["audio"]
 }

package/src/tasks/text-to-video/inference.ts

@@ -0,0 +1,57 @@
+/**
+ * Inference code generated from the JSON schema spec in ./spec
+ *
+ * Using src/scripts/inference-codegen
+ */
+/**
+ * Inputs for Text To Video inference
+ */
+export interface TextToVideoInput {
+	/**
+	 * The input text data (sometimes called "prompt")
+	 */
+	inputs: string;
+	/**
+	 * Additional inference parameters for Text To Video
+	 */
+	parameters?: TextToVideoParameters;
+	[property: string]: unknown;
+}
+/**
+ * Additional inference parameters for Text To Video
+ */
+export interface TextToVideoParameters {
+	/**
+	 * 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.
+	 */
+	guidance_scale?: number;
+	/**
+	 * One or several prompt to guide what NOT to include in image generation.
+	 */
+	negative_prompt?: string[];
+	/**
+	 * The num_frames parameter determines how many video frames are generated.
+	 */
+	num_frames?: number;
+	/**
+	 * The number of denoising steps. More denoising steps usually lead to a higher quality
+	 * image at the expense of slower inference.
+	 */
+	num_inference_steps?: number;
+	/**
+	 * Seed for the random number generator.
+	 */
+	seed?: number;
+	[property: string]: unknown;
+}
+/**
+ * Outputs of inference for the Text To Video task
+ */
+export interface TextToVideoOutput {
+	/**
+	 * The generated video returned as raw bytes in the payload.
+	 */
+	video: unknown;
+	[property: string]: unknown;
+}

package/src/tasks/text-to-video/spec/input.json

@@ -0,0 +1,49 @@
+{
+	"$id": "/inference/schemas/text-to-video/input.json",
+	"$schema": "http://json-schema.org/draft-06/schema#",
+	"description": "Inputs for Text To Video inference",
+	"title": "TextToVideoInput",
+	"type": "object",
+	"properties": {
+		"inputs": {
+			"description": "The input text data (sometimes called \"prompt\")",
+			"type": "string"
+		},
+		"parameters": {
+			"description": "Additional inference parameters for Text To Video",
+			"$ref": "#/$defs/TextToVideoParameters"
+		}
+	},
+	"$defs": {
+		"TextToVideoParameters": {
+			"title": "TextToVideoParameters",
+			"type": "object",
+			"properties": {
+				"num_frames": {
+					"type": "number",
+					"description": "The num_frames parameter determines how many video frames are generated."
+				},
+				"guidance_scale": {
+					"type": "number",
+					"description": "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": {
+					"type": "array",
+					"items": {
+						"type": "string"
+					},
+					"description": "One or several prompt to guide what NOT to include in image generation."
+				},
+				"num_inference_steps": {
+					"type": "integer",
+					"description": "The number of denoising steps. More denoising steps usually lead to a higher quality image at the expense of slower inference."
+				},
+				"seed": {
+					"type": "integer",
+					"description": "Seed for the random number generator."
+				}
+			}
+		}
+	},
+	"required": ["inputs"]
+}

package/src/tasks/text-to-video/spec/output.json

@@ -0,0 +1,13 @@
+{
+	"$id": "/inference/schemas/text-to-video/output.json",
+	"$schema": "http://json-schema.org/draft-06/schema#",
+	"description": "Outputs of inference for the Text To Video task",
+	"title": "TextToVideoOutput",
+	"type": "object",
+	"properties": {
+		"video": {
+			"description": "The generated video returned as raw bytes in the payload."
+		}
+	},
+	"required": ["video"]
+}

package/src/tasks/text2text-generation/inference.ts

@@ -3,7 +3,6 @@
  *
  * Using src/scripts/inference-codegen
  */
-
 /**
  * Inputs for Text2text Generation inference
  */
@@ -18,7 +17,6 @@ export interface Text2TextGenerationInput {
 	parameters?: Text2TextGenerationParameters;
 	[property: string]: unknown;
 }
-
 /**
  * Additional inference parameters for Text2text Generation
  */
@@ -30,16 +28,16 @@ export interface Text2TextGenerationParameters {
 	/**
 	 * Additional parametrization of the text generation algorithm
 	 */
-	generate_parameters?: { [key: string]: unknown };
+	generate_parameters?: {
+		[key: string]: unknown;
+	};
 	/**
 	 * The truncation strategy to use
 	 */
 	truncation?: Text2TextGenerationTruncationStrategy;
 	[property: string]: unknown;
 }
-
 export type Text2TextGenerationTruncationStrategy = "do_not_truncate" | "longest_first" | "only_first" | "only_second";
-
 /**
  * Outputs of inference for the Text2text Generation task
  */

package/src/tasks/translation/inference.ts

@@ -3,7 +3,6 @@
  *
  * Using src/scripts/inference-codegen
  */
-
 /**
  * Inputs for Translation inference
  */
@@ -18,7 +17,6 @@ export interface TranslationInput {
 	parameters?: TranslationParameters;
 	[property: string]: unknown;
 }
-
 /**
  * Additional inference parameters for Translation
  */
@@ -30,7 +28,9 @@ export interface TranslationParameters {
 	/**
 	 * Additional parametrization of the text generation algorithm.
 	 */
-	generate_parameters?: { [key: string]: unknown };
+	generate_parameters?: {
+		[key: string]: unknown;
+	};
 	/**
 	 * The source language of the text. Required for models that can translate from multiple
 	 * languages.
@@ -47,12 +47,10 @@ export interface TranslationParameters {
 	truncation?: TranslationTruncationStrategy;
 	[property: string]: unknown;
 }
-
 /**
  * The truncation strategy to use.
  */
 export type TranslationTruncationStrategy = "do_not_truncate" | "longest_first" | "only_first" | "only_second";
-
 /**
  * Outputs of inference for the Translation task
  */

package/src/tasks/visual-question-answering/inference.ts

@@ -28,7 +28,7 @@ export interface VisualQuestionAnsweringInputData {
 	/**
 	 * The question to answer based on the image.
 	 */
-	question: unknown;
+	question: string;
 	[property: string]: unknown;
 }
 /**

package/src/tasks/visual-question-answering/spec/input.json

@@ -11,10 +11,12 @@
 			"title": "VisualQuestionAnsweringInputData",
 			"properties": {
 				"image": {
-					"description": "The image."
+					"description": "The image.",
+					"comment": "type=binary"
 				},
 				"question": {
-					"description": "The question to answer based on the image."
+					"description": "The question to answer based on the image.",
+					"type": "string"
 				}
 			},
 			"required": ["question", "image"]

package/src/tasks/zero-shot-image-classification/inference.ts

@@ -10,7 +10,7 @@ export interface ZeroShotImageClassificationInput {
 	/**
 	 * The input image data to classify as a base64-encoded string.
 	 */
-	inputs: string;
+	inputs: Blob;
 	/**
 	 * Additional inference parameters for Zero Shot Image Classification
 	 */

package/src/tasks/zero-shot-image-classification/spec/input.json

@@ -7,7 +7,8 @@
 	"properties": {
 		"inputs": {
 			"type": "string",
-			"description": "The input image data to classify as a base64-encoded string."
+			"description": "The input image data to classify as a base64-encoded string.",
+			"comment": "type=binary"
 		},
 		"parameters": {
 			"description": "Additional inference parameters for Zero Shot Image Classification",

package/src/tasks/zero-shot-object-detection/inference.ts

@@ -10,7 +10,7 @@ export interface ZeroShotObjectDetectionInput {
 	/**
 	 * The input image data as a base64-encoded string.
 	 */
-	inputs: string;
+	inputs: Blob;
 	/**
 	 * Additional inference parameters for Zero Shot Object Detection
 	 */

package/src/tasks/zero-shot-object-detection/spec/input.json

@@ -7,7 +7,8 @@
 	"properties": {
 		"inputs": {
 			"description": "The input image data as a base64-encoded string.",
-			"type": "string"
+			"type": "string",
+			"comment": "type=binary"
 		},
 		"parameters": {
 			"description": "Additional inference parameters for Zero Shot Object Detection",