@huggingface/tasks 0.6.0 → 0.7.0

package/dist/index.cjs

@@ -23,7 +23,7 @@ __export(src_exports, {
   ALL_DISPLAY_MODEL_LIBRARY_KEYS: () => ALL_DISPLAY_MODEL_LIBRARY_KEYS,
   ALL_MODEL_LIBRARY_KEYS: () => ALL_MODEL_LIBRARY_KEYS,
   InferenceDisplayability: () => InferenceDisplayability,
-  LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS: () => LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS,
+  LIBRARY_TASK_MAPPING: () => LIBRARY_TASK_MAPPING,
   MAPPING_DEFAULT_WIDGET: () => MAPPING_DEFAULT_WIDGET,
   MODALITIES: () => MODALITIES,
   MODALITY_LABELS: () => MODALITY_LABELS,
@@ -40,7 +40,7 @@ __export(src_exports, {
 module.exports = __toCommonJS(src_exports);
 
 // src/library-to-tasks.ts
-var LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS = {
+var LIBRARY_TASK_MAPPING = {
   "adapter-transformers": ["question-answering", "text-classification", "token-classification"],
   allennlp: ["question-answering"],
   asteroid: [
@@ -76,6 +76,24 @@ var LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS = {
   ],
   stanza: ["token-classification"],
   timm: ["image-classification"],
+  transformers: [
+    "audio-classification",
+    "automatic-speech-recognition",
+    "depth-estimation",
+    "document-question-answering",
+    "fill-mask",
+    "image-classification",
+    "image-segmentation",
+    "image-to-text",
+    "image-to-image",
+    "object-detection",
+    "question-answering",
+    "text-generation",
+    "text2text-generation",
+    "visual-question-answering",
+    "zero-shot-classification",
+    "zero-shot-image-classification"
+  ],
   mindspore: ["image-classification"]
 };
 
@@ -3847,7 +3865,7 @@ var data_default34 = taskData34;
 // src/tasks/index.ts
 var TASKS_MODEL_LIBRARIES = {
   "audio-classification": ["speechbrain", "transformers", "transformers.js"],
-  "audio-to-audio": ["asteroid", "speechbrain"],
+  "audio-to-audio": ["asteroid", "fairseq", "speechbrain"],
   "automatic-speech-recognition": ["espnet", "nemo", "speechbrain", "transformers", "transformers.js"],
   "depth-estimation": ["transformers", "transformers.js"],
   "document-question-answering": ["transformers", "transformers.js"],
@@ -4084,7 +4102,7 @@ var flair = (model) => [
 tagger = SequenceTagger.load("${model.id}")`
 ];
 var gliner = (model) => [
-  `from model import GLiNER
+  `from gliner import GLiNER
 
 model = GLiNER.from_pretrained("${model.id}")`
 ];
@@ -4434,6 +4452,11 @@ var mlx = (model) => [
 export HF_HUB_ENABLE_HF_TRANS: string[]FER=1
 huggingface-cli download --local-dir ${nameWithoutNamespace(model.id)} ${model.id}`
 ];
+var mlxim = (model) => [
+  `from mlxim.model import create_model
+
+model = create_model(${model.id})`
+];
 var nemo = (model) => {
   let command = void 0;
   if (model.tags?.includes("automatic-speech-recognition")) {
@@ -4593,6 +4616,15 @@ var MODEL_LIBRARIES_UI_ELEMENTS = {
       term: { path: "gliner_config.json" }
     }
   },
+  grok: {
+    prettyLabel: "Grok",
+    repoName: "Grok",
+    repoUrl: "https://github.com/xai-org/grok-1",
+    filter: false,
+    countDownloads: {
+      terms: { path: ["ckpt/tensor00000_000", "ckpt-0/tensor00000_000"] }
+    }
+  },
   keras: {
     prettyLabel: "Keras",
     repoName: "Keras",
@@ -4628,6 +4660,15 @@ var MODEL_LIBRARIES_UI_ELEMENTS = {
     snippets: mlx,
     filter: true
   },
+  "mlx-image": {
+    prettyLabel: "mlx-image",
+    repoName: "mlx-image",
+    repoUrl: "https://github.com/riccardomusmeci/mlx-image",
+    docsUrl: "https://huggingface.co/docs/hub/mlx-image",
+    snippets: mlxim,
+    filter: false,
+    countDownloads: { term: { path: "model.safetensors" } }
+  },
   nemo: {
     prettyLabel: "NeMo",
     repoName: "NeMo",
@@ -5301,7 +5342,7 @@ function hasJsInferenceSnippet(model) {
   ALL_DISPLAY_MODEL_LIBRARY_KEYS,
   ALL_MODEL_LIBRARY_KEYS,
   InferenceDisplayability,
-  LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS,
+  LIBRARY_TASK_MAPPING,
   MAPPING_DEFAULT_WIDGET,
   MODALITIES,
   MODALITY_LABELS,

package/dist/index.d.ts

@@ -851,6 +851,17 @@ declare const MODEL_LIBRARIES_UI_ELEMENTS: {
             };
         };
     };
+    grok: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        filter: false;
+        countDownloads: {
+            terms: {
+                path: string[];
+            };
+        };
+    };
     keras: {
         prettyLabel: string;
         repoName: string;
@@ -894,6 +905,19 @@ declare const MODEL_LIBRARIES_UI_ELEMENTS: {
         snippets: (model: ModelData) => string[];
         filter: true;
     };
+    "mlx-image": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: false;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
     nemo: {
         prettyLabel: string;
         repoName: string;
@@ -1117,17 +1141,17 @@ declare const MODEL_LIBRARIES_UI_ELEMENTS: {
     };
 };
 type ModelLibraryKey = keyof typeof MODEL_LIBRARIES_UI_ELEMENTS;
-declare const ALL_MODEL_LIBRARY_KEYS: ("sklearn" | "adapter-transformers" | "allennlp" | "asteroid" | "audiocraft" | "bertopic" | "diffusers" | "doctr" | "espnet" | "fairseq" | "fastai" | "fasttext" | "flair" | "gliner" | "keras" | "k2" | "mindspore" | "ml-agents" | "mlx" | "nemo" | "open_clip" | "paddlenlp" | "peft" | "pyannote-audio" | "pythae" | "sample-factory" | "sentence-transformers" | "setfit" | "spacy" | "span-marker" | "speechbrain" | "stable-baselines3" | "stanza" | "tensorflowtts" | "timm" | "transformers" | "transformers.js" | "unity-sentis" | "whisperkit")[];
-declare const ALL_DISPLAY_MODEL_LIBRARY_KEYS: ("sklearn" | "adapter-transformers" | "allennlp" | "asteroid" | "audiocraft" | "bertopic" | "diffusers" | "doctr" | "espnet" | "fairseq" | "fastai" | "fasttext" | "flair" | "gliner" | "keras" | "k2" | "mindspore" | "ml-agents" | "mlx" | "nemo" | "open_clip" | "paddlenlp" | "peft" | "pyannote-audio" | "pythae" | "sample-factory" | "sentence-transformers" | "setfit" | "spacy" | "span-marker" | "speechbrain" | "stable-baselines3" | "stanza" | "tensorflowtts" | "timm" | "transformers" | "transformers.js" | "unity-sentis" | "whisperkit")[];
+declare const ALL_MODEL_LIBRARY_KEYS: ("sklearn" | "adapter-transformers" | "allennlp" | "asteroid" | "audiocraft" | "bertopic" | "diffusers" | "doctr" | "espnet" | "fairseq" | "fastai" | "fasttext" | "flair" | "gliner" | "grok" | "keras" | "k2" | "mindspore" | "ml-agents" | "mlx" | "mlx-image" | "nemo" | "open_clip" | "paddlenlp" | "peft" | "pyannote-audio" | "pythae" | "sample-factory" | "sentence-transformers" | "setfit" | "spacy" | "span-marker" | "speechbrain" | "stable-baselines3" | "stanza" | "tensorflowtts" | "timm" | "transformers" | "transformers.js" | "unity-sentis" | "whisperkit")[];
+declare const ALL_DISPLAY_MODEL_LIBRARY_KEYS: ("sklearn" | "adapter-transformers" | "allennlp" | "asteroid" | "audiocraft" | "bertopic" | "diffusers" | "doctr" | "espnet" | "fairseq" | "fastai" | "fasttext" | "flair" | "gliner" | "grok" | "keras" | "k2" | "mindspore" | "ml-agents" | "mlx" | "mlx-image" | "nemo" | "open_clip" | "paddlenlp" | "peft" | "pyannote-audio" | "pythae" | "sample-factory" | "sentence-transformers" | "setfit" | "spacy" | "span-marker" | "speechbrain" | "stable-baselines3" | "stanza" | "tensorflowtts" | "timm" | "transformers" | "transformers.js" | "unity-sentis" | "whisperkit")[];
 
 /**
- * Mapping from library name (excluding Transformers) to its supported tasks.
+ * Mapping from library name to its supported tasks.
  * Inference API (serverless) should be disabled for all other (library, task) pairs beyond this mapping.
- * As an exception, we assume Transformers supports all inference tasks.
- * This mapping is generated automatically by "python-api-export-tasks" action in huggingface/api-inference-community repo upon merge.
- * Ref: https://github.com/huggingface/api-inference-community/pull/158
+ * This mapping is partially generated automatically by "python-api-export-tasks" action in
+ * huggingface/api-inference-community repo upon merge. For transformers, the mapping is manually
+ * based on api-inference.
  */
-declare const LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS: Partial<Record<ModelLibraryKey, PipelineType[]>>;
+declare const LIBRARY_TASK_MAPPING: Partial<Record<ModelLibraryKey, PipelineType[]>>;
 
 type PerLanguageMapping = Map<WidgetType, string[] | WidgetExample[]>;
 declare const MAPPING_DEFAULT_WIDGET: Map<string, PerLanguageMapping>;
@@ -1338,6 +1362,154 @@ interface AutomaticSpeechRecognitionOutputChunk {
     [property: string]: unknown;
 }
 
+/**
+ * Inference code generated from the JSON schema spec in ./spec
+ *
+ * Using src/scripts/inference-codegen
+ */
+/**
+ * Inputs for ChatCompletion inference
+ */
+interface ChatCompletionInput {
+    /**
+     * Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing
+     * frequency in the text so far, decreasing the model's likelihood to repeat the same line
+     * verbatim.
+     */
+    frequency_penalty?: number;
+    /**
+     * The maximum number of tokens that can be generated in the chat completion.
+     */
+    max_tokens?: number;
+    messages: ChatCompletionInputMessage[];
+    /**
+     * The random sampling seed.
+     */
+    seed?: number;
+    /**
+     * Stop generating tokens if a stop token is generated.
+     */
+    stop?: ChatCompletionInputStopReason;
+    /**
+     * If set, partial message deltas will be sent.
+     */
+    stream?: boolean;
+    /**
+     * The value used to modulate the logits distribution.
+     */
+    temperature?: number;
+    /**
+     * If set to < 1, only the smallest set of most probable tokens with probabilities that add
+     * up to `top_p` or higher are kept for generation.
+     */
+    top_p?: number;
+    [property: string]: unknown;
+}
+interface ChatCompletionInputMessage {
+    /**
+     * The content of the message.
+     */
+    content: string;
+    role: ChatCompletionMessageRole;
+    [property: string]: unknown;
+}
+/**
+ * The role of the message author.
+ */
+type ChatCompletionMessageRole = "assistant" | "system" | "user";
+/**
+ * Stop generating tokens if a stop token is generated.
+ */
+type ChatCompletionInputStopReason = string[] | string;
+/**
+ * Outputs for Chat Completion inference
+ */
+interface ChatCompletionOutput {
+    /**
+     * A list of chat completion choices.
+     */
+    choices: ChatCompletionOutputChoice[];
+    /**
+     * The Unix timestamp (in seconds) of when the chat completion was created.
+     */
+    created: number;
+    [property: string]: unknown;
+}
+interface ChatCompletionOutputChoice {
+    /**
+     * The reason why the generation was stopped.
+     */
+    finish_reason: ChatCompletionFinishReason;
+    /**
+     * The index of the choice in the list of choices.
+     */
+    index: number;
+    message: ChatCompletionOutputChoiceMessage;
+    [property: string]: unknown;
+}
+/**
+ * The reason why the generation was stopped.
+ *
+ * The generated sequence reached the maximum allowed length
+ *
+ * The model generated an end-of-sentence (EOS) token
+ *
+ * One of the sequence in stop_sequences was generated
+ */
+type ChatCompletionFinishReason = "length" | "eos_token" | "stop_sequence";
+interface ChatCompletionOutputChoiceMessage {
+    /**
+     * The content of the chat completion message.
+     */
+    content: string;
+    role: ChatCompletionMessageRole;
+    [property: string]: unknown;
+}
+/**
+ * Chat Completion Stream Output
+ */
+interface ChatCompletionStreamOutput {
+    /**
+     * A list of chat completion choices.
+     */
+    choices: ChatCompletionStreamOutputChoice[];
+    /**
+     * The Unix timestamp (in seconds) of when the chat completion was created. Each chunk has
+     * the same timestamp.
+     */
+    created: number;
+    [property: string]: unknown;
+}
+interface ChatCompletionStreamOutputChoice {
+    /**
+     * A chat completion delta generated by streamed model responses.
+     */
+    delta: ChatCompletionStreamOutputDelta;
+    /**
+     * The reason why the generation was stopped.
+     */
+    finish_reason?: ChatCompletionFinishReason;
+    /**
+     * The index of the choice in the list of choices.
+     */
+    index: number;
+    [property: string]: unknown;
+}
+/**
+ * A chat completion delta generated by streamed model responses.
+ */
+interface ChatCompletionStreamOutputDelta {
+    /**
+     * The contents of the chunk message.
+     */
+    content?: string;
+    /**
+     * The role of the author of this message.
+     */
+    role?: string;
+    [property: string]: unknown;
+}
+
 /**
  * Inference code generated from the JSON schema spec in ./spec
  *
@@ -2622,6 +2794,10 @@ interface TextGenerationInput {
      * Additional inference parameters
      */
     parameters?: TextGenerationParameters;
+    /**
+     * Whether to stream output tokens
+     */
+    stream?: boolean;
     [property: string]: unknown;
 }
 /**
@@ -2717,16 +2893,16 @@ interface TextGenerationOutputDetails {
     /**
      * Details about additional sequences when best_of is provided
      */
-    best_of_sequences?: TextGenerationSequenceDetails[];
+    best_of_sequences?: TextGenerationOutputSequenceDetails[];
     /**
      * The reason why the generation was stopped.
      */
-    finish_reason: FinishReason;
+    finish_reason: TextGenerationFinishReason;
     /**
      * The number of generated tokens
      */
     generated_tokens: number;
-    prefill: PrefillToken[];
+    prefill: TextGenerationPrefillToken[];
     /**
      * The random seed used for generation
      */
@@ -2734,23 +2910,24 @@ interface TextGenerationOutputDetails {
     /**
      * The generated tokens and associated details
      */
-    tokens: Token[];
-    [property: string]: unknown;
-}
-interface TextGenerationSequenceDetails {
+    tokens: TextGenerationOutputToken[];
     /**
-     * The reason why the generation was stopped.
+     * Most likely tokens
      */
-    finish_reason: FinishReason;
+    top_tokens?: Array<TextGenerationOutputToken[]>;
+    [property: string]: unknown;
+}
+interface TextGenerationOutputSequenceDetails {
+    finish_reason: TextGenerationFinishReason;
     /**
      * The generated text
      */
-    generated_text: number;
+    generated_text: string;
     /**
      * The number of generated tokens
      */
     generated_tokens: number;
-    prefill: PrefillToken[];
+    prefill: TextGenerationPrefillToken[];
     /**
      * The random seed used for generation
      */
@@ -2758,18 +2935,24 @@ interface TextGenerationSequenceDetails {
     /**
      * The generated tokens and associated details
      */
-    tokens: Token[];
+    tokens: TextGenerationOutputToken[];
+    /**
+     * Most likely tokens
+     */
+    top_tokens?: Array<TextGenerationOutputToken[]>;
     [property: string]: unknown;
 }
 /**
- * The generated sequence reached the maximum allowed length
+ * The reason why the generation was stopped.
  *
- * The model generated an end-of-sentence (EOS) token
+ * length: The generated sequence reached the maximum allowed length
  *
- * One of the sequence in stop_sequences was generated
+ * eos_token: The model generated an end-of-sentence (EOS) token
+ *
+ * stop_sequence: One of the sequence in stop_sequences was generated
  */
-type FinishReason = "length" | "eos_token" | "stop_sequence";
-interface PrefillToken {
+type TextGenerationFinishReason = "length" | "eos_token" | "stop_sequence";
+interface TextGenerationPrefillToken {
     id: number;
     logprob: number;
     /**
@@ -2778,9 +2961,12 @@ interface PrefillToken {
     text: string;
     [property: string]: unknown;
 }
-interface Token {
+/**
+ * Generated token.
+ */
+interface TextGenerationOutputToken {
     id: number;
-    logprob: number;
+    logprob?: number;
     /**
      * Whether or not that token is a special one
      */
@@ -3285,4 +3471,4 @@ declare namespace index {
   };
 }
 
-export { ALL_DISPLAY_MODEL_LIBRARY_KEYS, ALL_MODEL_LIBRARY_KEYS, AddedToken, AudioClassificationInput, AudioClassificationOutput, AudioClassificationOutputElement, AudioClassificationParameters, AutomaticSpeechRecognitionInput, AutomaticSpeechRecognitionOutput, AutomaticSpeechRecognitionOutputChunk, AutomaticSpeechRecognitionParameters, BoundingBox, ChatMessage, ClassificationOutputTransform$1 as ClassificationOutputTransform, DepthEstimationInput, DepthEstimationOutput, DocumentQuestionAnsweringInput, DocumentQuestionAnsweringInputData, DocumentQuestionAnsweringOutput, DocumentQuestionAnsweringOutputElement, DocumentQuestionAnsweringParameters, EarlyStoppingUnion$2 as EarlyStoppingUnion, ExampleRepo, FeatureExtractionInput, FeatureExtractionOutput, FillMaskInput, FillMaskOutput, FillMaskOutputElement, FillMaskParameters, FinishReason, GenerationParameters$2 as GenerationParameters, ImageClassificationInput, ImageClassificationOutput, ImageClassificationOutputElement, ImageClassificationParameters, ImageSegmentationInput, ImageSegmentationOutput, ImageSegmentationOutputElement, ImageSegmentationParameters, ImageSegmentationSubtask, ImageToImageInput, ImageToImageOutput, ImageToImageParameters, ImageToTextInput, ImageToTextOutput, ImageToTextParameters, InferenceDisplayability, LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS, LibraryUiElement, MAPPING_DEFAULT_WIDGET, MODALITIES, MODALITY_LABELS, MODEL_LIBRARIES_UI_ELEMENTS, Modality, ModelData, ModelLibraryKey, ObjectDetectionInput, ObjectDetectionOutput, ObjectDetectionOutputElement, ObjectDetectionParameters, PIPELINE_DATA, PIPELINE_TYPES, PIPELINE_TYPES_SET, PipelineData, PipelineType, PrefillToken, QuestionAnsweringInput, QuestionAnsweringInputData, QuestionAnsweringOutput, QuestionAnsweringOutputElement, QuestionAnsweringParameters, SPECIAL_TOKENS_ATTRIBUTES, SUBTASK_TYPES, SentenceSimilarityInput, SentenceSimilarityInputData, SentenceSimilarityOutput, SpecialTokensMap, SummarizationInput, SummarizationOutput, TASKS_DATA, TASKS_MODEL_LIBRARIES, TableQuestionAnsweringInput, TableQuestionAnsweringInputData, TableQuestionAnsweringOutput, TableQuestionAnsweringOutputElement, TargetSize$1 as TargetSize, TaskData, TaskDataCustom, TaskDemo, TaskDemoEntry, Text2TextGenerationParameters, Text2TextGenerationTruncationStrategy, TextClassificationInput, TextClassificationOutput, TextClassificationOutputElement, TextClassificationParameters, TextGenerationInput, TextGenerationOutput, TextGenerationOutputDetails, TextGenerationParameters, TextGenerationSequenceDetails, TextToAudioParameters, TextToImageInput, TextToImageOutput, TextToImageParameters, TextToSpeechInput, TextToSpeechOutput, Token, TokenClassificationAggregationStrategy, TokenClassificationInput, TokenClassificationOutput, TokenClassificationOutputElement, TokenClassificationParameters, TokenizerConfig, TransformersInfo, TranslationInput, TranslationOutput, VideoClassificationInput, VideoClassificationOutput, VideoClassificationOutputElement, VideoClassificationParameters, VisualQuestionAnsweringInput, VisualQuestionAnsweringInputData, VisualQuestionAnsweringOutput, VisualQuestionAnsweringOutputElement, VisualQuestionAnsweringParameters, WidgetExample, WidgetExampleAssetAndPromptInput, WidgetExampleAssetAndTextInput, WidgetExampleAssetAndZeroShotInput, WidgetExampleAssetInput, WidgetExampleAttribute, WidgetExampleChatInput, WidgetExampleOutput, WidgetExampleOutputAnswerScore, WidgetExampleOutputLabels, WidgetExampleOutputText, WidgetExampleOutputUrl, WidgetExampleSentenceSimilarityInput, WidgetExampleStructuredDataInput, WidgetExampleTableDataInput, WidgetExampleTextAndContextInput, WidgetExampleTextAndTableInput, WidgetExampleTextInput, WidgetExampleZeroShotTextInput, WidgetType, WordBox, ZeroShotClassificationInput, ZeroShotClassificationInputData, ZeroShotClassificationOutput, ZeroShotClassificationOutputElement, ZeroShotClassificationParameters, ZeroShotImageClassificationInput, ZeroShotImageClassificationInputData, ZeroShotImageClassificationOutput, ZeroShotImageClassificationOutputElement, ZeroShotImageClassificationParameters, ZeroShotObjectDetectionInput, ZeroShotObjectDetectionInputData, ZeroShotObjectDetectionOutput, ZeroShotObjectDetectionOutputElement, index as snippets };
+export { ALL_DISPLAY_MODEL_LIBRARY_KEYS, ALL_MODEL_LIBRARY_KEYS, AddedToken, AudioClassificationInput, AudioClassificationOutput, AudioClassificationOutputElement, AudioClassificationParameters, AutomaticSpeechRecognitionInput, AutomaticSpeechRecognitionOutput, AutomaticSpeechRecognitionOutputChunk, AutomaticSpeechRecognitionParameters, BoundingBox, ChatCompletionFinishReason, ChatCompletionInput, ChatCompletionInputMessage, ChatCompletionOutput, ChatCompletionOutputChoice, ChatCompletionOutputChoiceMessage, ChatCompletionStreamOutput, ChatCompletionStreamOutputChoice, ChatCompletionStreamOutputDelta, ChatMessage, ClassificationOutputTransform$1 as ClassificationOutputTransform, DepthEstimationInput, DepthEstimationOutput, DocumentQuestionAnsweringInput, DocumentQuestionAnsweringInputData, DocumentQuestionAnsweringOutput, DocumentQuestionAnsweringOutputElement, DocumentQuestionAnsweringParameters, EarlyStoppingUnion$2 as EarlyStoppingUnion, ExampleRepo, FeatureExtractionInput, FeatureExtractionOutput, FillMaskInput, FillMaskOutput, FillMaskOutputElement, FillMaskParameters, GenerationParameters$2 as GenerationParameters, ImageClassificationInput, ImageClassificationOutput, ImageClassificationOutputElement, ImageClassificationParameters, ImageSegmentationInput, ImageSegmentationOutput, ImageSegmentationOutputElement, ImageSegmentationParameters, ImageSegmentationSubtask, ImageToImageInput, ImageToImageOutput, ImageToImageParameters, ImageToTextInput, ImageToTextOutput, ImageToTextParameters, InferenceDisplayability, LIBRARY_TASK_MAPPING, LibraryUiElement, MAPPING_DEFAULT_WIDGET, MODALITIES, MODALITY_LABELS, MODEL_LIBRARIES_UI_ELEMENTS, Modality, ModelData, ModelLibraryKey, ObjectDetectionInput, ObjectDetectionOutput, ObjectDetectionOutputElement, ObjectDetectionParameters, PIPELINE_DATA, PIPELINE_TYPES, PIPELINE_TYPES_SET, PipelineData, PipelineType, QuestionAnsweringInput, QuestionAnsweringInputData, QuestionAnsweringOutput, QuestionAnsweringOutputElement, QuestionAnsweringParameters, SPECIAL_TOKENS_ATTRIBUTES, SUBTASK_TYPES, SentenceSimilarityInput, SentenceSimilarityInputData, SentenceSimilarityOutput, SpecialTokensMap, SummarizationInput, SummarizationOutput, TASKS_DATA, TASKS_MODEL_LIBRARIES, TableQuestionAnsweringInput, TableQuestionAnsweringInputData, TableQuestionAnsweringOutput, TableQuestionAnsweringOutputElement, TargetSize$1 as TargetSize, TaskData, TaskDataCustom, TaskDemo, TaskDemoEntry, Text2TextGenerationParameters, Text2TextGenerationTruncationStrategy, TextClassificationInput, TextClassificationOutput, TextClassificationOutputElement, TextClassificationParameters, TextGenerationFinishReason, TextGenerationInput, TextGenerationOutput, TextGenerationOutputDetails, TextGenerationOutputSequenceDetails, TextGenerationOutputToken, TextGenerationParameters, TextGenerationPrefillToken, TextToAudioParameters, TextToImageInput, TextToImageOutput, TextToImageParameters, TextToSpeechInput, TextToSpeechOutput, TokenClassificationAggregationStrategy, TokenClassificationInput, TokenClassificationOutput, TokenClassificationOutputElement, TokenClassificationParameters, TokenizerConfig, TransformersInfo, TranslationInput, TranslationOutput, VideoClassificationInput, VideoClassificationOutput, VideoClassificationOutputElement, VideoClassificationParameters, VisualQuestionAnsweringInput, VisualQuestionAnsweringInputData, VisualQuestionAnsweringOutput, VisualQuestionAnsweringOutputElement, VisualQuestionAnsweringParameters, WidgetExample, WidgetExampleAssetAndPromptInput, WidgetExampleAssetAndTextInput, WidgetExampleAssetAndZeroShotInput, WidgetExampleAssetInput, WidgetExampleAttribute, WidgetExampleChatInput, WidgetExampleOutput, WidgetExampleOutputAnswerScore, WidgetExampleOutputLabels, WidgetExampleOutputText, WidgetExampleOutputUrl, WidgetExampleSentenceSimilarityInput, WidgetExampleStructuredDataInput, WidgetExampleTableDataInput, WidgetExampleTextAndContextInput, WidgetExampleTextAndTableInput, WidgetExampleTextInput, WidgetExampleZeroShotTextInput, WidgetType, WordBox, ZeroShotClassificationInput, ZeroShotClassificationInputData, ZeroShotClassificationOutput, ZeroShotClassificationOutputElement, ZeroShotClassificationParameters, ZeroShotImageClassificationInput, ZeroShotImageClassificationInputData, ZeroShotImageClassificationOutput, ZeroShotImageClassificationOutputElement, ZeroShotImageClassificationParameters, ZeroShotObjectDetectionInput, ZeroShotObjectDetectionInputData, ZeroShotObjectDetectionOutput, ZeroShotObjectDetectionOutputElement, index as snippets };

package/dist/index.js

@@ -5,7 +5,7 @@ var __export = (target, all) => {
 };
 
 // src/library-to-tasks.ts
-var LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS = {
+var LIBRARY_TASK_MAPPING = {
   "adapter-transformers": ["question-answering", "text-classification", "token-classification"],
   allennlp: ["question-answering"],
   asteroid: [
@@ -41,6 +41,24 @@ var LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS = {
   ],
   stanza: ["token-classification"],
   timm: ["image-classification"],
+  transformers: [
+    "audio-classification",
+    "automatic-speech-recognition",
+    "depth-estimation",
+    "document-question-answering",
+    "fill-mask",
+    "image-classification",
+    "image-segmentation",
+    "image-to-text",
+    "image-to-image",
+    "object-detection",
+    "question-answering",
+    "text-generation",
+    "text2text-generation",
+    "visual-question-answering",
+    "zero-shot-classification",
+    "zero-shot-image-classification"
+  ],
   mindspore: ["image-classification"]
 };
 
@@ -3812,7 +3830,7 @@ var data_default34 = taskData34;
 // src/tasks/index.ts
 var TASKS_MODEL_LIBRARIES = {
   "audio-classification": ["speechbrain", "transformers", "transformers.js"],
-  "audio-to-audio": ["asteroid", "speechbrain"],
+  "audio-to-audio": ["asteroid", "fairseq", "speechbrain"],
   "automatic-speech-recognition": ["espnet", "nemo", "speechbrain", "transformers", "transformers.js"],
   "depth-estimation": ["transformers", "transformers.js"],
   "document-question-answering": ["transformers", "transformers.js"],
@@ -4049,7 +4067,7 @@ var flair = (model) => [
 tagger = SequenceTagger.load("${model.id}")`
 ];
 var gliner = (model) => [
-  `from model import GLiNER
+  `from gliner import GLiNER
 
 model = GLiNER.from_pretrained("${model.id}")`
 ];
@@ -4399,6 +4417,11 @@ var mlx = (model) => [
 export HF_HUB_ENABLE_HF_TRANS: string[]FER=1
 huggingface-cli download --local-dir ${nameWithoutNamespace(model.id)} ${model.id}`
 ];
+var mlxim = (model) => [
+  `from mlxim.model import create_model
+
+model = create_model(${model.id})`
+];
 var nemo = (model) => {
   let command = void 0;
   if (model.tags?.includes("automatic-speech-recognition")) {
@@ -4558,6 +4581,15 @@ var MODEL_LIBRARIES_UI_ELEMENTS = {
       term: { path: "gliner_config.json" }
     }
   },
+  grok: {
+    prettyLabel: "Grok",
+    repoName: "Grok",
+    repoUrl: "https://github.com/xai-org/grok-1",
+    filter: false,
+    countDownloads: {
+      terms: { path: ["ckpt/tensor00000_000", "ckpt-0/tensor00000_000"] }
+    }
+  },
   keras: {
     prettyLabel: "Keras",
     repoName: "Keras",
@@ -4593,6 +4625,15 @@ var MODEL_LIBRARIES_UI_ELEMENTS = {
     snippets: mlx,
     filter: true
   },
+  "mlx-image": {
+    prettyLabel: "mlx-image",
+    repoName: "mlx-image",
+    repoUrl: "https://github.com/riccardomusmeci/mlx-image",
+    docsUrl: "https://huggingface.co/docs/hub/mlx-image",
+    snippets: mlxim,
+    filter: false,
+    countDownloads: { term: { path: "model.safetensors" } }
+  },
   nemo: {
     prettyLabel: "NeMo",
     repoName: "NeMo",
@@ -5265,7 +5306,7 @@ export {
   ALL_DISPLAY_MODEL_LIBRARY_KEYS,
   ALL_MODEL_LIBRARY_KEYS,
   InferenceDisplayability,
-  LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS,
+  LIBRARY_TASK_MAPPING,
   MAPPING_DEFAULT_WIDGET,
   MODALITIES,
   MODALITY_LABELS,

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@huggingface/tasks",
   "packageManager": "pnpm@8.10.5",
-  "version": "0.6.0",
+  "version": "0.7.0",
   "description": "List of ML tasks for huggingface.co/tasks",
   "repository": "https://github.com/huggingface/huggingface.js.git",
   "publishConfig": {
@@ -33,7 +33,7 @@
   "license": "MIT",
   "devDependencies": {
     "@types/node": "^20.11.5",
-    "quicktype-core": "https://github.com/huggingface/quicktype/raw/pack-18.0.15/packages/quicktype-core/quicktype-core-18.0.15.tgz"
+    "quicktype-core": "https://github.com/huggingface/quicktype/raw/pack-18.0.17/packages/quicktype-core/quicktype-core-18.0.17.tgz"
   },
   "scripts": {
     "lint": "eslint --quiet --fix --ext .cjs,.ts .",

package/src/index.ts

@@ -1,4 +1,4 @@
-export { LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS } from "./library-to-tasks";
+export { LIBRARY_TASK_MAPPING } from "./library-to-tasks";
 export { MAPPING_DEFAULT_WIDGET } from "./default-widget-inputs";
 export type { TaskData, TaskDemo, TaskDemoEntry, ExampleRepo } from "./tasks";
 export * from "./tasks";

package/src/library-to-tasks.ts

@@ -2,13 +2,13 @@ import type { ModelLibraryKey } from "./model-libraries";
 import type { PipelineType } from "./pipelines";
 
 /**
- * Mapping from library name (excluding Transformers) to its supported tasks.
+ * Mapping from library name to its supported tasks.
  * Inference API (serverless) should be disabled for all other (library, task) pairs beyond this mapping.
- * As an exception, we assume Transformers supports all inference tasks.
- * This mapping is generated automatically by "python-api-export-tasks" action in huggingface/api-inference-community repo upon merge.
- * Ref: https://github.com/huggingface/api-inference-community/pull/158
+ * This mapping is partially generated automatically by "python-api-export-tasks" action in
+ * huggingface/api-inference-community repo upon merge. For transformers, the mapping is manually
+ * based on api-inference.
  */
-export const LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS: Partial<Record<ModelLibraryKey, PipelineType[]>> = {
+export const LIBRARY_TASK_MAPPING: Partial<Record<ModelLibraryKey, PipelineType[]>> = {
 	"adapter-transformers": ["question-answering", "text-classification", "token-classification"],
 	allennlp: ["question-answering"],
 	asteroid: [
@@ -44,5 +44,23 @@ export const LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS: Partial<Record<ModelLi
 	],
 	stanza: ["token-classification"],
 	timm: ["image-classification"],
+	transformers: [
+		"audio-classification",
+		"automatic-speech-recognition",
+		"depth-estimation",
+		"document-question-answering",
+		"fill-mask",
+		"image-classification",
+		"image-segmentation",
+		"image-to-text",
+		"image-to-image",
+		"object-detection",
+		"question-answering",
+		"text-generation",
+		"text2text-generation",
+		"visual-question-answering",
+		"zero-shot-classification",
+		"zero-shot-image-classification",
+	],
 	mindspore: ["image-classification"],
 };

package/src/model-libraries-snippets.ts

@@ -141,7 +141,7 @@ tagger = SequenceTagger.load("${model.id}")`,
 ];
 
 export const gliner = (model: ModelData): string[] => [
-	`from model import GLiNER
+	`from gliner import GLiNER
 
 model = GLiNER.from_pretrained("${model.id}")`,
 ];
@@ -538,6 +538,12 @@ export HF_HUB_ENABLE_HF_TRANS: string[]FER=1
 huggingface-cli download --local-dir ${nameWithoutNamespace(model.id)} ${model.id}`,
 ];
 
+export const mlxim = (model: ModelData): string[] => [
+	`from mlxim.model import create_model
+
+model = create_model(${model.id})`,
+];
+
 export const nemo = (model: ModelData): string[] => {
 	let command: string[] | undefined = undefined;
 	// Resolve the tag to a nemo domain/sub-domain