@huggingface/tasks 0.2.1 → 0.3.0

package/dist/index.d.ts

@@ -1,50 +1,3 @@
-/**
- * Add your new library here.
- *
- * This is for modeling (= architectures) libraries, not for file formats (like ONNX, etc).
- * File formats live in an enum inside the internal codebase.
- */
-declare enum ModelLibrary {
-    "adapter-transformers" = "Adapter Transformers",
-    "allennlp" = "allenNLP",
-    "asteroid" = "Asteroid",
-    "bertopic" = "BERTopic",
-    "diffusers" = "Diffusers",
-    "doctr" = "docTR",
-    "espnet" = "ESPnet",
-    "fairseq" = "Fairseq",
-    "flair" = "Flair",
-    "keras" = "Keras",
-    "k2" = "K2",
-    "mlx" = "MLX",
-    "nemo" = "NeMo",
-    "open_clip" = "OpenCLIP",
-    "paddlenlp" = "PaddleNLP",
-    "peft" = "PEFT",
-    "pyannote-audio" = "pyannote.audio",
-    "sample-factory" = "Sample Factory",
-    "sentence-transformers" = "Sentence Transformers",
-    "setfit" = "SetFit",
-    "sklearn" = "Scikit-learn",
-    "spacy" = "spaCy",
-    "span-marker" = "SpanMarker",
-    "speechbrain" = "speechbrain",
-    "tensorflowtts" = "TensorFlowTTS",
-    "timm" = "Timm",
-    "fastai" = "fastai",
-    "transformers" = "Transformers",
-    "transformers.js" = "Transformers.js",
-    "stanza" = "Stanza",
-    "fasttext" = "fastText",
-    "stable-baselines3" = "Stable-Baselines3",
-    "ml-agents" = "Unity ML-Agents",
-    "pythae" = "Pythae",
-    "mindspore" = "MindSpore",
-    "unity-sentis" = "Unity Sentis"
-}
-type ModelLibraryKey = keyof typeof ModelLibrary;
-declare const ALL_DISPLAY_MODEL_LIBRARY_KEYS: ("speechbrain" | "fastai" | "adapter-transformers" | "allennlp" | "asteroid" | "bertopic" | "diffusers" | "doctr" | "espnet" | "fairseq" | "flair" | "keras" | "k2" | "mlx" | "nemo" | "open_clip" | "paddlenlp" | "peft" | "pyannote-audio" | "sample-factory" | "sentence-transformers" | "setfit" | "sklearn" | "spacy" | "span-marker" | "tensorflowtts" | "timm" | "transformers" | "transformers.js" | "stanza" | "fasttext" | "stable-baselines3" | "ml-agents" | "pythae" | "mindspore" | "unity-sentis")[];
-
 declare const MODALITIES: readonly ["cv", "nlp", "audio", "tabular", "multimodal", "rl", "other"];
 type Modality = (typeof MODALITIES)[number];
 declare const MODALITY_LABELS: {
@@ -446,15 +399,6 @@ declare const PIPELINE_TYPES: ("other" | "text-classification" | "token-classifi
 declare const SUBTASK_TYPES: string[];
 declare const PIPELINE_TYPES_SET: Set<"other" | "text-classification" | "token-classification" | "table-question-answering" | "question-answering" | "zero-shot-classification" | "translation" | "summarization" | "conversational" | "feature-extraction" | "text-generation" | "text2text-generation" | "fill-mask" | "sentence-similarity" | "text-to-speech" | "text-to-audio" | "automatic-speech-recognition" | "audio-to-audio" | "audio-classification" | "voice-activity-detection" | "depth-estimation" | "image-classification" | "object-detection" | "image-segmentation" | "text-to-image" | "image-to-text" | "image-to-image" | "image-to-video" | "unconditional-image-generation" | "video-classification" | "reinforcement-learning" | "robotics" | "tabular-classification" | "tabular-regression" | "tabular-to-text" | "table-to-text" | "multiple-choice" | "text-retrieval" | "time-series-forecasting" | "text-to-video" | "visual-question-answering" | "document-question-answering" | "zero-shot-image-classification" | "graph-ml" | "mask-generation" | "zero-shot-object-detection" | "text-to-3d" | "image-to-3d">;
 
-/**
- * Mapping from library name (excluding Transformers) to its supported tasks.
- * Inference API 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
- */
-declare const LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS: Partial<Record<ModelLibraryKey, PipelineType[]>>;
-
 /**
  * See default-widget-inputs.ts for the default widget inputs, this files only contains the types
  */
@@ -612,7 +556,7 @@ interface ModelData {
      */
     widgetData?: WidgetExample[] | undefined;
     /**
-     * Parameters that will be used by the widget when calling Inference API
+     * Parameters that will be used by the widget when calling Inference Endpoints (serverless)
      * https://huggingface.co/docs/api-inference/detailed_parameters
      *
      * can be set in the model card metadata (under `inference/parameters`)
@@ -655,21 +599,48 @@ interface TransformersInfo {
     processor?: string;
 }
 
+/**
+ * This file contains the (simplified) types used
+ * to represent queries that are made to Elastic
+ * in order to count number of model downloads
+ *
+ * Read this doc about download stats on the Hub:
+ *
+ * https://huggingface.co/docs/hub/models-download-stats
+ *
+ * see also:
+ * https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-bool-query.html
+ */
+type ElasticBoolQueryFilter = {
+    term?: {
+        path: string;
+    };
+} | {
+    terms?: {
+        path: string[];
+    };
+} | {
+    wildcard?: {
+        path: string;
+    };
+};
+
 /**
  * Elements configurable by a model library.
  */
 interface LibraryUiElement {
     /**
-     * Name displayed on the main
+     * Pretty name of the library.
+     * displayed in tags, and on the main
      * call-to-action button on the model page.
      */
-    btnLabel: string;
+    prettyLabel: string;
     /**
-     * Repo name
+     * Repo name of the library's (usually on GitHub) code repo
      */
     repoName: string;
     /**
-     * URL to library's repo
+     * URL to library's (usually on GitHub) code repo
      */
     repoUrl: string;
     /**
@@ -677,11 +648,401 @@ interface LibraryUiElement {
      */
     docsUrl?: string;
     /**
-     * Code snippet displayed on model page
+     * Code snippet(s) displayed on model page
+     */
+    snippets?: (model: ModelData) => string[];
+    /**
+     * Elastic query used to count this library's model downloads
+     *
+     * By default, those files are counted:
+     * "config.json", "config.yaml", "hyperparams.yaml", "meta.yaml"
      */
-    snippets: (model: ModelData) => string[];
+    countDownloads?: ElasticBoolQueryFilter;
+    /**
+     * should we display this library in hf.co/models filter
+     * (only for popular libraries with > 100 models)
+     */
+    filter?: boolean;
 }
-declare const MODEL_LIBRARIES_UI_ELEMENTS: Partial<Record<ModelLibraryKey, LibraryUiElement>>;
+/**
+ * Add your new library here.
+ *
+ * This is for modeling (= architectures) libraries, not for file formats (like ONNX, etc).
+ * (unlike libraries, file formats live in an enum inside the internal codebase.)
+ *
+ * Doc on how to add a library to the Hub:
+ *
+ * https://huggingface.co/docs/hub/models-adding-libraries
+ *
+ * /!\ IMPORTANT
+ *
+ * The key you choose is the tag your models have in their library_name on the Hub.
+ */
+declare const MODEL_LIBRARIES_UI_ELEMENTS: {
+    "adapter-transformers": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    allennlp: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    asteroid: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    bertopic: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    diffusers: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    doctr: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+    };
+    espnet: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    fairseq: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    fastai: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    fasttext: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    flair: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    keras: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    k2: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+    };
+    mindspore: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+    };
+    "ml-agents": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            wildcard: {
+                path: string;
+            };
+        };
+    };
+    mlx: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    nemo: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            wildcard: {
+                path: string;
+            };
+        };
+    };
+    open_clip: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            wildcard: {
+                path: string;
+            };
+        };
+    };
+    paddlenlp: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    peft: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    "pyannote-audio": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    pythae: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    "sample-factory": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    "sentence-transformers": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    setfit: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    sklearn: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    spacy: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            wildcard: {
+                path: string;
+            };
+        };
+    };
+    "span-marker": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    speechbrain: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    "stable-baselines3": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            wildcard: {
+                path: string;
+            };
+        };
+    };
+    stanza: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            term: {
+                path: string;
+            };
+        };
+    };
+    tensorflowtts: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: (model: ModelData) => string[];
+    };
+    timm: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+        countDownloads: {
+            terms: {
+                path: string[];
+            };
+        };
+    };
+    transformers: {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    "transformers.js": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        docsUrl: string;
+        snippets: (model: ModelData) => string[];
+        filter: true;
+    };
+    "unity-sentis": {
+        prettyLabel: string;
+        repoName: string;
+        repoUrl: string;
+        snippets: () => string[];
+        filter: true;
+        countDownloads: {
+            wildcard: {
+                path: string;
+            };
+        };
+    };
+};
+type ModelLibraryKey = keyof typeof MODEL_LIBRARIES_UI_ELEMENTS;
+declare const ALL_MODEL_LIBRARY_KEYS: ("sklearn" | "adapter-transformers" | "allennlp" | "asteroid" | "bertopic" | "diffusers" | "doctr" | "espnet" | "fairseq" | "fastai" | "fasttext" | "flair" | "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")[];
+declare const ALL_DISPLAY_MODEL_LIBRARY_KEYS: ("sklearn" | "adapter-transformers" | "allennlp" | "asteroid" | "bertopic" | "diffusers" | "doctr" | "espnet" | "fairseq" | "fastai" | "fasttext" | "flair" | "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")[];
+
+/**
+ * Mapping from library name (excluding Transformers) 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
+ */
+declare const LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS: Partial<Record<ModelLibraryKey, PipelineType[]>>;
 
 type PerLanguageMapping = Map<PipelineType, string[] | WidgetExample[]>;
 declare const MAPPING_DEFAULT_WIDGET: Map<string, PerLanguageMapping>;
@@ -731,6 +1092,7 @@ interface TaskData {
     datasets: ExampleRepo[];
     demo: TaskDemo;
     id: PipelineType;
+    canonicalId?: PipelineType;
     isPlaceholder?: boolean;
     label: string;
     libraries: ModelLibraryKey[];
@@ -850,4 +1212,4 @@ declare namespace index {
   };
 }
 
-export { ALL_DISPLAY_MODEL_LIBRARY_KEYS, ExampleRepo, InferenceDisplayability, LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS, LibraryUiElement, MAPPING_DEFAULT_WIDGET, MODALITIES, MODALITY_LABELS, MODEL_LIBRARIES_UI_ELEMENTS, Modality, ModelData, ModelLibrary, ModelLibraryKey, PIPELINE_DATA, PIPELINE_TYPES, PIPELINE_TYPES_SET, PipelineData, PipelineType, SUBTASK_TYPES, TASKS_DATA, TASKS_MODEL_LIBRARIES, TaskData, TaskDataCustom, TaskDemo, TaskDemoEntry, TransformersInfo, WidgetExample, WidgetExampleAssetAndPromptInput, WidgetExampleAssetAndTextInput, WidgetExampleAssetAndZeroShotInput, WidgetExampleAssetInput, WidgetExampleAttribute, WidgetExampleOutput, WidgetExampleOutputAnswerScore, WidgetExampleOutputLabels, WidgetExampleOutputText, WidgetExampleOutputUrl, WidgetExampleSentenceSimilarityInput, WidgetExampleStructuredDataInput, WidgetExampleTableDataInput, WidgetExampleTextAndContextInput, WidgetExampleTextAndTableInput, WidgetExampleTextInput, WidgetExampleZeroShotTextInput, index as snippets };
+export { ALL_DISPLAY_MODEL_LIBRARY_KEYS, ALL_MODEL_LIBRARY_KEYS, ExampleRepo, InferenceDisplayability, LIBRARY_TASK_MAPPING_EXCLUDING_TRANSFORMERS, LibraryUiElement, MAPPING_DEFAULT_WIDGET, MODALITIES, MODALITY_LABELS, MODEL_LIBRARIES_UI_ELEMENTS, Modality, ModelData, ModelLibraryKey, PIPELINE_DATA, PIPELINE_TYPES, PIPELINE_TYPES_SET, PipelineData, PipelineType, SUBTASK_TYPES, TASKS_DATA, TASKS_MODEL_LIBRARIES, TaskData, TaskDataCustom, TaskDemo, TaskDemoEntry, TransformersInfo, WidgetExample, WidgetExampleAssetAndPromptInput, WidgetExampleAssetAndTextInput, WidgetExampleAssetAndZeroShotInput, WidgetExampleAssetInput, WidgetExampleAttribute, WidgetExampleOutput, WidgetExampleOutputAnswerScore, WidgetExampleOutputLabels, WidgetExampleOutputText, WidgetExampleOutputUrl, WidgetExampleSentenceSimilarityInput, WidgetExampleStructuredDataInput, WidgetExampleTableDataInput, WidgetExampleTextAndContextInput, WidgetExampleTextAndTableInput, WidgetExampleTextInput, WidgetExampleZeroShotTextInput, index as snippets };