@huggingface/tasks 0.13.2 → 0.13.3

package/dist/commonjs/index.d.ts

@@ -12,6 +12,7 @@ export { SPECIAL_TOKENS_ATTRIBUTES } from "./tokenizer-data.js";
 import * as snippets from "./snippets/index.js";
 export * from "./gguf.js";
 export { snippets };
+export type { InferenceSnippet } from "./snippets/index.js";
 export { SKUS, DEFAULT_MEMORY_OPTIONS } from "./hardware.js";
 export type { HardwareSpec, SkuType } from "./hardware.js";
 export { LOCAL_APPS } from "./local-apps.js";

package/dist/commonjs/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,oBAAoB,EAAE,MAAM,uBAAuB,CAAC;AAC7D,OAAO,EAAE,sBAAsB,EAAE,MAAM,4BAA4B,CAAC;AACpE,YAAY,EAAE,QAAQ,EAAE,QAAQ,EAAE,aAAa,EAAE,WAAW,EAAE,MAAM,kBAAkB,CAAC;AACvF,cAAc,kBAAkB,CAAC;AACjC,OAAO,EACN,aAAa,EACb,cAAc,EACd,KAAK,UAAU,EACf,KAAK,YAAY,EACjB,KAAK,YAAY,EACjB,KAAK,QAAQ,EACb,UAAU,EACV,eAAe,EACf,aAAa,EACb,kBAAkB,GAClB,MAAM,gBAAgB,CAAC;AACxB,OAAO,EACN,8BAA8B,EAC9B,sBAAsB,EACtB,2BAA2B,GAC3B,MAAM,sBAAsB,CAAC;AAC9B,YAAY,EAAE,gBAAgB,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAC9E,YAAY,EAAE,SAAS,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AACnE,YAAY,EAAE,UAAU,EAAE,gBAAgB,EAAE,eAAe,EAAE,MAAM,qBAAqB,CAAC;AACzF,YAAY,EACX,aAAa,EACb,sBAAsB,EACtB,gCAAgC,EAChC,8BAA8B,EAC9B,kCAAkC,EAClC,uBAAuB,EACvB,sBAAsB,EACtB,oCAAoC,EACpC,gCAAgC,EAChC,2BAA2B,EAC3B,gCAAgC,EAChC,8BAA8B,EAC9B,sBAAsB,EACtB,8BAA8B,EAC9B,mBAAmB,EACnB,sBAAsB,EACtB,yBAAyB,EACzB,8BAA8B,EAC9B,uBAAuB,GACvB,MAAM,qBAAqB,CAAC;AAC7B,OAAO,EAAE,yBAAyB,EAAE,MAAM,qBAAqB,CAAC;AAEhE,OAAO,KAAK,QAAQ,MAAM,qBAAqB,CAAC;AAChD,cAAc,WAAW,CAAC;AAE1B,OAAO,EAAE,QAAQ,EAAE,CAAC;AAEpB,OAAO,EAAE,IAAI,EAAE,sBAAsB,EAAE,MAAM,eAAe,CAAC;AAC7D,YAAY,EAAE,YAAY,EAAE,OAAO,EAAE,MAAM,eAAe,CAAC;AAC3D,OAAO,EAAE,UAAU,EAAE,MAAM,iBAAiB,CAAC;AAC7C,YAAY,EAAE,QAAQ,EAAE,WAAW,EAAE,eAAe,EAAE,MAAM,iBAAiB,CAAC;AAE9E,OAAO,EAAE,6BAA6B,EAAE,MAAM,wBAAwB,CAAC;AACvE,YAAY,EAAE,uBAAuB,EAAE,iBAAiB,EAAE,MAAM,wBAAwB,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,oBAAoB,EAAE,MAAM,uBAAuB,CAAC;AAC7D,OAAO,EAAE,sBAAsB,EAAE,MAAM,4BAA4B,CAAC;AACpE,YAAY,EAAE,QAAQ,EAAE,QAAQ,EAAE,aAAa,EAAE,WAAW,EAAE,MAAM,kBAAkB,CAAC;AACvF,cAAc,kBAAkB,CAAC;AACjC,OAAO,EACN,aAAa,EACb,cAAc,EACd,KAAK,UAAU,EACf,KAAK,YAAY,EACjB,KAAK,YAAY,EACjB,KAAK,QAAQ,EACb,UAAU,EACV,eAAe,EACf,aAAa,EACb,kBAAkB,GAClB,MAAM,gBAAgB,CAAC;AACxB,OAAO,EACN,8BAA8B,EAC9B,sBAAsB,EACtB,2BAA2B,GAC3B,MAAM,sBAAsB,CAAC;AAC9B,YAAY,EAAE,gBAAgB,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAC9E,YAAY,EAAE,SAAS,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AACnE,YAAY,EAAE,UAAU,EAAE,gBAAgB,EAAE,eAAe,EAAE,MAAM,qBAAqB,CAAC;AACzF,YAAY,EACX,aAAa,EACb,sBAAsB,EACtB,gCAAgC,EAChC,8BAA8B,EAC9B,kCAAkC,EAClC,uBAAuB,EACvB,sBAAsB,EACtB,oCAAoC,EACpC,gCAAgC,EAChC,2BAA2B,EAC3B,gCAAgC,EAChC,8BAA8B,EAC9B,sBAAsB,EACtB,8BAA8B,EAC9B,mBAAmB,EACnB,sBAAsB,EACtB,yBAAyB,EACzB,8BAA8B,EAC9B,uBAAuB,GACvB,MAAM,qBAAqB,CAAC;AAC7B,OAAO,EAAE,yBAAyB,EAAE,MAAM,qBAAqB,CAAC;AAEhE,OAAO,KAAK,QAAQ,MAAM,qBAAqB,CAAC;AAChD,cAAc,WAAW,CAAC;AAE1B,OAAO,EAAE,QAAQ,EAAE,CAAC;AACpB,YAAY,EAAE,gBAAgB,EAAE,MAAM,qBAAqB,CAAC;AAG5D,OAAO,EAAE,IAAI,EAAE,sBAAsB,EAAE,MAAM,eAAe,CAAC;AAC7D,YAAY,EAAE,YAAY,EAAE,OAAO,EAAE,MAAM,eAAe,CAAC;AAC3D,OAAO,EAAE,UAAU,EAAE,MAAM,iBAAiB,CAAC;AAC7C,YAAY,EAAE,QAAQ,EAAE,WAAW,EAAE,eAAe,EAAE,MAAM,iBAAiB,CAAC;AAE9E,OAAO,EAAE,6BAA6B,EAAE,MAAM,wBAAwB,CAAC;AACvE,YAAY,EAAE,uBAAuB,EAAE,iBAAiB,EAAE,MAAM,wBAAwB,CAAC"}

package/dist/commonjs/pipelines.d.ts

@@ -170,6 +170,12 @@ export declare const PIPELINE_DATA: {
         modality: "audio";
         color: "green";
     };
+    "audio-text-to-text": {
+        name: string;
+        modality: "multimodal";
+        color: "red";
+        hideInDatasets: true;
+    };
     "voice-activity-detection": {
         name: string;
         modality: "audio";
@@ -421,5 +427,5 @@ export type PipelineType = keyof typeof PIPELINE_DATA;
 export type WidgetType = PipelineType | "conversational";
 export declare const PIPELINE_TYPES: PipelineType[];
 export declare const SUBTASK_TYPES: string[];
-export declare const PIPELINE_TYPES_SET: Set<"other" | "text-classification" | "token-classification" | "table-question-answering" | "question-answering" | "zero-shot-classification" | "translation" | "summarization" | "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" | "image-text-to-text" | "visual-question-answering" | "document-question-answering" | "zero-shot-image-classification" | "graph-ml" | "mask-generation" | "zero-shot-object-detection" | "text-to-3d" | "image-to-3d" | "image-feature-extraction" | "video-text-to-text" | "keypoint-detection" | "any-to-any">;
+export declare const PIPELINE_TYPES_SET: Set<"other" | "text-classification" | "token-classification" | "table-question-answering" | "question-answering" | "zero-shot-classification" | "translation" | "summarization" | "feature-extraction" | "text-generation" | "text2text-generation" | "fill-mask" | "sentence-similarity" | "text-to-speech" | "text-to-audio" | "automatic-speech-recognition" | "audio-to-audio" | "audio-classification" | "audio-text-to-text" | "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" | "image-text-to-text" | "visual-question-answering" | "document-question-answering" | "zero-shot-image-classification" | "graph-ml" | "mask-generation" | "zero-shot-object-detection" | "text-to-3d" | "image-to-3d" | "image-feature-extraction" | "video-text-to-text" | "keypoint-detection" | "any-to-any">;
 //# sourceMappingURL=pipelines.d.ts.map

package/dist/commonjs/pipelines.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"pipelines.d.ts","sourceRoot":"","sources":["../../src/pipelines.ts"],"names":[],"mappings":"AAAA,eAAO,MAAM,UAAU,yEAA0E,CAAC;AAElG,MAAM,MAAM,QAAQ,GAAG,CAAC,OAAO,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC;AAEnD,eAAO,MAAM,eAAe;;;;;;;;CAQQ,CAAC;AAErC;;;;;;GAMG;AACH,MAAM,WAAW,OAAO;IACvB;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;IACb;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;CACb;AAED;;;;;GAKG;AACH,MAAM,WAAW,YAAY;IAC5B;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,CAAC,EAAE,OAAO,EAAE,CAAC;IACrB,QAAQ,EAAE,QAAQ,CAAC;IACnB;;OAEG;IACH,KAAK,EAAE,MAAM,GAAG,OAAO,GAAG,QAAQ,GAAG,QAAQ,GAAG,KAAK,GAAG,QAAQ,CAAC;IACjE;;OAEG;IACH,YAAY,CAAC,EAAE,OAAO,CAAC;IACvB;;OAEG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;CACzB;AAcD,eAAO,MAAM,aAAa;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;CAsmBc,CAAC;AAEzC,MAAM,MAAM,YAAY,GAAG,MAAM,OAAO,aAAa,CAAC;AAEtD,MAAM,MAAM,UAAU,GAAG,YAAY,GAAG,gBAAgB,CAAC;AAEzD,eAAO,MAAM,cAAc,EAAiC,YAAY,EAAE,CAAC;AAE3E,eAAO,MAAM,aAAa,UAEN,CAAC;AAErB,eAAO,MAAM,kBAAkB,ipCAA0B,CAAC"}
+{"version":3,"file":"pipelines.d.ts","sourceRoot":"","sources":["../../src/pipelines.ts"],"names":[],"mappings":"AAAA,eAAO,MAAM,UAAU,yEAA0E,CAAC;AAElG,MAAM,MAAM,QAAQ,GAAG,CAAC,OAAO,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC;AAEnD,eAAO,MAAM,eAAe;;;;;;;;CAQQ,CAAC;AAErC;;;;;;GAMG;AACH,MAAM,WAAW,OAAO;IACvB;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;IACb;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;CACb;AAED;;;;;GAKG;AACH,MAAM,WAAW,YAAY;IAC5B;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,CAAC,EAAE,OAAO,EAAE,CAAC;IACrB,QAAQ,EAAE,QAAQ,CAAC;IACnB;;OAEG;IACH,KAAK,EAAE,MAAM,GAAG,OAAO,GAAG,QAAQ,GAAG,QAAQ,GAAG,KAAK,GAAG,QAAQ,CAAC;IACjE;;OAEG;IACH,YAAY,CAAC,EAAE,OAAO,CAAC;IACvB;;OAEG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;CACzB;AAcD,eAAO,MAAM,aAAa;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;CA4mBc,CAAC;AAEzC,MAAM,MAAM,YAAY,GAAG,MAAM,OAAO,aAAa,CAAC;AAEtD,MAAM,MAAM,UAAU,GAAG,YAAY,GAAG,gBAAgB,CAAC;AAEzD,eAAO,MAAM,cAAc,EAAiC,YAAY,EAAE,CAAC;AAE3E,eAAO,MAAM,aAAa,UAEN,CAAC;AAErB,eAAO,MAAM,kBAAkB,wqCAA0B,CAAC"}

package/dist/commonjs/pipelines.js

@@ -309,6 +309,12 @@ exports.PIPELINE_DATA = {
         modality: "audio",
         color: "green",
     },
+    "audio-text-to-text": {
+        name: "Audio-Text-to-Text",
+        modality: "multimodal",
+        color: "red",
+        hideInDatasets: true,
+    },
     "voice-activity-detection": {
         name: "Voice Activity Detection",
         modality: "audio",

package/dist/commonjs/tasks/automatic-speech-recognition/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/automatic-speech-recognition/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cA6Ef,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/automatic-speech-recognition/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAyFf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/commonjs/tasks/automatic-speech-recognition/data.js

@@ -6,12 +6,16 @@ const taskData = {
             description: "31,175 hours of multilingual audio-text dataset in 108 languages.",
             id: "mozilla-foundation/common_voice_17_0",
         },
+        {
+            description: "Multilingual and diverse audio dataset with 101k hours of audio.",
+            id: "amphion/Emilia-Dataset",
+        },
         {
             description: "A dataset with 44.6k hours of English speaker data and 6k hours of other language speakers.",
             id: "parler-tts/mls_eng",
         },
         {
-            description: "A multi-lingual audio dataset with 370K hours of audio.",
+            description: "A multilingual audio dataset with 370K hours of audio.",
             id: "espnet/yodas",
         },
     ],
@@ -54,6 +58,10 @@ const taskData = {
             description: "An end-to-end model that performs ASR and Speech Translation by MetaAI.",
             id: "facebook/seamless-m4t-v2-large",
         },
+        {
+            description: "A powerful multilingual ASR and Speech Translation model by Nvidia.",
+            id: "nvidia/canary-1b",
+        },
         {
             description: "Powerful speaker diarization model.",
             id: "pyannote/speaker-diarization-3.1",
@@ -65,13 +73,17 @@ const taskData = {
             id: "hf-audio/whisper-large-v3",
         },
         {
-            description: "Fastest speech recognition application.",
-            id: "sanchit-gandhi/whisper-jax",
+            description: "Latest ASR model from Useful Sensors.",
+            id: "mrfakename/Moonshinex",
         },
         {
             description: "A high quality speech and text translation model by Meta.",
             id: "facebook/seamless_m4t",
         },
+        {
+            description: "A powerful multilingual ASR and Speech Translation model by Nvidia",
+            id: "nvidia/canary-1b",
+        },
     ],
     summary: "Automatic Speech Recognition (ASR), also known as Speech to Text (STT), is the task of transcribing a given audio to text. It has many applications, such as voice user interfaces.",
     widgetModels: ["openai/whisper-large-v3"],

package/dist/commonjs/tasks/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/tasks/index.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAC;AA4CpD,mBAAmB,qCAAqC,CAAC;AACzD,mBAAmB,6CAA6C,CAAC;AACjE,YAAY,EACX,mBAAmB,EACnB,0BAA0B,EAC1B,oBAAoB,EACpB,4BAA4B,EAC5B,2BAA2B,EAC3B,0BAA0B,EAC1B,gCAAgC,EAChC,+BAA+B,GAC/B,MAAM,gCAAgC,CAAC;AACxC,mBAAmB,4CAA4C,CAAC;AAChE,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,0BAA0B,CAAC;AAC9C,YAAY,EACX,wBAAwB,EACxB,yBAAyB,EACzB,gCAAgC,EAChC,6BAA6B,GAC7B,MAAM,qCAAqC,CAAC;AAC7C,mBAAmB,+BAA+B,CAAC;AACnD,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,qBAAqB,EAAE,MAAM,8BAA8B,CAAC;AAC/G,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,iCAAiC,CAAC;AACrD,mBAAmB,iCAAiC,CAAC;AACrD,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,oCAAoC,CAAC;AACxD,mBAAmB,8BAA8B,CAAC;AAClD,mBAAmB,yCAAyC,CAAC;AAC7D,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,qBAAqB,EAAE,MAAM,8BAA8B,CAAC;AAC/G,YAAY,EAAE,sBAAsB,EAAE,iBAAiB,EAAE,kBAAkB,EAAE,MAAM,+BAA+B,CAAC;AACnH,mBAAmB,qCAAqC,CAAC;AACzD,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,MAAM,4BAA4B,CAAC;AACtF,YAAY,EACX,6BAA6B,EAC7B,uBAAuB,EACvB,wBAAwB,EACxB,+BAA+B,EAC/B,4BAA4B,GAC5B,MAAM,oCAAoC,CAAC;AAC5C,YAAY,EACX,gCAAgC,EAChC,gCAAgC,EAChC,mBAAmB,EACnB,oBAAoB,EACpB,2BAA2B,EAC3B,qCAAqC,EACrC,kCAAkC,EAClC,yBAAyB,EACzB,uCAAuC,EACvC,0BAA0B,GAC1B,MAAM,gCAAgC,CAAC;AACxC,mBAAmB,qCAAqC,CAAC;AACzD,mBAAmB,0CAA0C,CAAC;AAC9D,mBAAmB,yCAAyC,CAAC;AAC7D,mBAAmB,+CAA+C,CAAC;AACnE,YAAY,EACX,WAAW,EACX,4BAA4B,EAC5B,gCAAgC,EAChC,6BAA6B,EAC7B,oCAAoC,GACpC,MAAM,2CAA2C,CAAC;AAEnD,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,uBAAuB,CAAC;AAE7D;;GAEG;AACH,eAAO,MAAM,qBAAqB,EAAE,MAAM,CAAC,YAAY,EAAE,eAAe,EAAE,CA6DzE,CAAC;AAoBF,eAAO,MAAM,UAAU,EAAE,MAAM,CAAC,YAAY,EAAE,QAAQ,GAAG,SAAS,CAqDxD,CAAC;AAEX,MAAM,WAAW,WAAW;IAC3B,WAAW,EAAE,MAAM,CAAC;IACpB,EAAE,EAAE,MAAM,CAAC;CACX;AAED,MAAM,MAAM,aAAa,GACtB;IACA,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,OAAO,CAAC;CACb,GACD;IACA,IAAI,EAAE,KAAK,CAAC;QACX,KAAK,EAAE,MAAM,CAAC;QACd,KAAK,EAAE,MAAM,CAAC;KACd,CAAC,CAAC;IACH,IAAI,EAAE,OAAO,CAAC;CACb,GACD;IACA,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,KAAK,CAAC;CACX,GACD;IACA,KAAK,EAAE,MAAM,EAAE,EAAE,CAAC;IAClB,IAAI,EAAE,SAAS,CAAC;CACf,GACD;IACA,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACZ,GACD;IACA,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,KAAK,CAAC;QACb,GAAG,EAAE,MAAM,CAAC;QACZ,KAAK,EAAE,MAAM,CAAC;QACd,IAAI,EAAE,MAAM,CAAC;KACb,CAAC,CAAC;IACH,IAAI,EAAE,kBAAkB,CAAC;CACxB,CAAC;AAEL,MAAM,WAAW,QAAQ;IACxB,MAAM,EAAE,aAAa,EAAE,CAAC;IACxB,OAAO,EAAE,aAAa,EAAE,CAAC;CACzB;AAED,MAAM,WAAW,QAAQ;IACxB,QAAQ,EAAE,WAAW,EAAE,CAAC;IACxB,IAAI,EAAE,QAAQ,CAAC;IACf,EAAE,EAAE,YAAY,CAAC;IACjB,WAAW,CAAC,EAAE,YAAY,CAAC;IAC3B,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,KAAK,EAAE,MAAM,CAAC;IACd,SAAS,EAAE,eAAe,EAAE,CAAC;IAC7B,OAAO,EAAE,WAAW,EAAE,CAAC;IACvB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,OAAO,EAAE,MAAM,CAAC;IAChB,YAAY,EAAE,MAAM,EAAE,CAAC;IACvB,SAAS,CAAC,EAAE,MAAM,CAAC;CACnB;AAED,MAAM,MAAM,cAAc,GAAG,IAAI,CAAC,QAAQ,EAAE,IAAI,GAAG,OAAO,GAAG,WAAW,CAAC,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/tasks/index.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAC;AA4CpD,mBAAmB,qCAAqC,CAAC;AACzD,mBAAmB,6CAA6C,CAAC;AACjE,YAAY,EACX,mBAAmB,EACnB,0BAA0B,EAC1B,oBAAoB,EACpB,4BAA4B,EAC5B,2BAA2B,EAC3B,0BAA0B,EAC1B,gCAAgC,EAChC,+BAA+B,GAC/B,MAAM,gCAAgC,CAAC;AACxC,mBAAmB,4CAA4C,CAAC;AAChE,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,0BAA0B,CAAC;AAC9C,YAAY,EACX,wBAAwB,EACxB,yBAAyB,EACzB,gCAAgC,EAChC,6BAA6B,GAC7B,MAAM,qCAAqC,CAAC;AAC7C,mBAAmB,+BAA+B,CAAC;AACnD,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,qBAAqB,EAAE,MAAM,8BAA8B,CAAC;AAC/G,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,iCAAiC,CAAC;AACrD,mBAAmB,iCAAiC,CAAC;AACrD,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,oCAAoC,CAAC;AACxD,mBAAmB,8BAA8B,CAAC;AAClD,mBAAmB,yCAAyC,CAAC;AAC7D,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,qBAAqB,EAAE,MAAM,8BAA8B,CAAC;AAC/G,YAAY,EAAE,sBAAsB,EAAE,iBAAiB,EAAE,kBAAkB,EAAE,MAAM,+BAA+B,CAAC;AACnH,mBAAmB,qCAAqC,CAAC;AACzD,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,MAAM,4BAA4B,CAAC;AACtF,YAAY,EACX,6BAA6B,EAC7B,uBAAuB,EACvB,wBAAwB,EACxB,+BAA+B,EAC/B,4BAA4B,GAC5B,MAAM,oCAAoC,CAAC;AAC5C,YAAY,EACX,gCAAgC,EAChC,gCAAgC,EAChC,mBAAmB,EACnB,oBAAoB,EACpB,2BAA2B,EAC3B,qCAAqC,EACrC,kCAAkC,EAClC,yBAAyB,EACzB,uCAAuC,EACvC,0BAA0B,GAC1B,MAAM,gCAAgC,CAAC;AACxC,mBAAmB,qCAAqC,CAAC;AACzD,mBAAmB,0CAA0C,CAAC;AAC9D,mBAAmB,yCAAyC,CAAC;AAC7D,mBAAmB,+CAA+C,CAAC;AACnE,YAAY,EACX,WAAW,EACX,4BAA4B,EAC5B,gCAAgC,EAChC,6BAA6B,EAC7B,oCAAoC,GACpC,MAAM,2CAA2C,CAAC;AAEnD,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,uBAAuB,CAAC;AAE7D;;GAEG;AACH,eAAO,MAAM,qBAAqB,EAAE,MAAM,CAAC,YAAY,EAAE,eAAe,EAAE,CA8DzE,CAAC;AAoBF,eAAO,MAAM,UAAU,EAAE,MAAM,CAAC,YAAY,EAAE,QAAQ,GAAG,SAAS,CAsDxD,CAAC;AAEX,MAAM,WAAW,WAAW;IAC3B,WAAW,EAAE,MAAM,CAAC;IACpB,EAAE,EAAE,MAAM,CAAC;CACX;AAED,MAAM,MAAM,aAAa,GACtB;IACA,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,OAAO,CAAC;CACb,GACD;IACA,IAAI,EAAE,KAAK,CAAC;QACX,KAAK,EAAE,MAAM,CAAC;QACd,KAAK,EAAE,MAAM,CAAC;KACd,CAAC,CAAC;IACH,IAAI,EAAE,OAAO,CAAC;CACb,GACD;IACA,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,KAAK,CAAC;CACX,GACD;IACA,KAAK,EAAE,MAAM,EAAE,EAAE,CAAC;IAClB,IAAI,EAAE,SAAS,CAAC;CACf,GACD;IACA,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACZ,GACD;IACA,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,KAAK,CAAC;QACb,GAAG,EAAE,MAAM,CAAC;QACZ,KAAK,EAAE,MAAM,CAAC;QACd,IAAI,EAAE,MAAM,CAAC;KACb,CAAC,CAAC;IACH,IAAI,EAAE,kBAAkB,CAAC;CACxB,CAAC;AAEL,MAAM,WAAW,QAAQ;IACxB,MAAM,EAAE,aAAa,EAAE,CAAC;IACxB,OAAO,EAAE,aAAa,EAAE,CAAC;CACzB;AAED,MAAM,WAAW,QAAQ;IACxB,QAAQ,EAAE,WAAW,EAAE,CAAC;IACxB,IAAI,EAAE,QAAQ,CAAC;IACf,EAAE,EAAE,YAAY,CAAC;IACjB,WAAW,CAAC,EAAE,YAAY,CAAC;IAC3B,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,KAAK,EAAE,MAAM,CAAC;IACd,SAAS,EAAE,eAAe,EAAE,CAAC;IAC7B,OAAO,EAAE,WAAW,EAAE,CAAC;IACvB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,OAAO,EAAE,MAAM,CAAC;IAChB,YAAY,EAAE,MAAM,EAAE,CAAC;IACvB,SAAS,CAAC,EAAE,MAAM,CAAC;CACnB;AAED,MAAM,MAAM,cAAc,GAAG,IAAI,CAAC,QAAQ,EAAE,IAAI,GAAG,OAAO,GAAG,WAAW,CAAC,CAAC"}

package/dist/commonjs/tasks/index.js

@@ -52,6 +52,7 @@ exports.TASKS_MODEL_LIBRARIES = {
     "audio-classification": ["speechbrain", "transformers", "transformers.js"],
     "audio-to-audio": ["asteroid", "fairseq", "speechbrain"],
     "automatic-speech-recognition": ["espnet", "nemo", "speechbrain", "transformers", "transformers.js"],
+    "audio-text-to-text": [],
     "depth-estimation": ["transformers", "transformers.js"],
     "document-question-answering": ["transformers", "transformers.js"],
     "feature-extraction": ["sentence-transformers", "transformers", "transformers.js"],
@@ -131,6 +132,7 @@ exports.TASKS_DATA = {
     "any-to-any": getData("any-to-any", data_js_16.default),
     "audio-classification": getData("audio-classification", data_js_1.default),
     "audio-to-audio": getData("audio-to-audio", data_js_2.default),
+    "audio-text-to-text": getData("audio-text-to-text", data_js_16.default),
     "automatic-speech-recognition": getData("automatic-speech-recognition", data_js_3.default),
     "depth-estimation": getData("depth-estimation", data_js_15.default),
     "document-question-answering": getData("document-question-answering", data_js_4.default),

package/dist/commonjs/tasks/mask-generation/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/mask-generation/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAkDf,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/mask-generation/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAgEf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/commonjs/tasks/mask-generation/data.js

@@ -1,7 +1,16 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
 const taskData = {
-    datasets: [],
+    datasets: [
+        {
+            description: "Widely used benchmark dataset for multiple Vision tasks.",
+            id: "merve/coco2017",
+        },
+        {
+            description: "Medical Imaging dataset of the Human Brain for segmentation and mask generating tasks",
+            id: "rocky93/BraTS_segmentation",
+        },
+    ],
     demo: {
         inputs: [
             {
@@ -16,7 +25,12 @@ const taskData = {
             },
         ],
     },
-    metrics: [],
+    metrics: [
+        {
+            description: "IoU is used to measure the overlap between predicted mask and the ground truth mask.",
+            id: "Intersection over Union (IoU)",
+        },
+    ],
     models: [
         {
             description: "Small yet powerful mask generation model.",

package/dist/esm/index.d.ts

@@ -12,6 +12,7 @@ export { SPECIAL_TOKENS_ATTRIBUTES } from "./tokenizer-data.js";
 import * as snippets from "./snippets/index.js";
 export * from "./gguf.js";
 export { snippets };
+export type { InferenceSnippet } from "./snippets/index.js";
 export { SKUS, DEFAULT_MEMORY_OPTIONS } from "./hardware.js";
 export type { HardwareSpec, SkuType } from "./hardware.js";
 export { LOCAL_APPS } from "./local-apps.js";

package/dist/esm/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,oBAAoB,EAAE,MAAM,uBAAuB,CAAC;AAC7D,OAAO,EAAE,sBAAsB,EAAE,MAAM,4BAA4B,CAAC;AACpE,YAAY,EAAE,QAAQ,EAAE,QAAQ,EAAE,aAAa,EAAE,WAAW,EAAE,MAAM,kBAAkB,CAAC;AACvF,cAAc,kBAAkB,CAAC;AACjC,OAAO,EACN,aAAa,EACb,cAAc,EACd,KAAK,UAAU,EACf,KAAK,YAAY,EACjB,KAAK,YAAY,EACjB,KAAK,QAAQ,EACb,UAAU,EACV,eAAe,EACf,aAAa,EACb,kBAAkB,GAClB,MAAM,gBAAgB,CAAC;AACxB,OAAO,EACN,8BAA8B,EAC9B,sBAAsB,EACtB,2BAA2B,GAC3B,MAAM,sBAAsB,CAAC;AAC9B,YAAY,EAAE,gBAAgB,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAC9E,YAAY,EAAE,SAAS,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AACnE,YAAY,EAAE,UAAU,EAAE,gBAAgB,EAAE,eAAe,EAAE,MAAM,qBAAqB,CAAC;AACzF,YAAY,EACX,aAAa,EACb,sBAAsB,EACtB,gCAAgC,EAChC,8BAA8B,EAC9B,kCAAkC,EAClC,uBAAuB,EACvB,sBAAsB,EACtB,oCAAoC,EACpC,gCAAgC,EAChC,2BAA2B,EAC3B,gCAAgC,EAChC,8BAA8B,EAC9B,sBAAsB,EACtB,8BAA8B,EAC9B,mBAAmB,EACnB,sBAAsB,EACtB,yBAAyB,EACzB,8BAA8B,EAC9B,uBAAuB,GACvB,MAAM,qBAAqB,CAAC;AAC7B,OAAO,EAAE,yBAAyB,EAAE,MAAM,qBAAqB,CAAC;AAEhE,OAAO,KAAK,QAAQ,MAAM,qBAAqB,CAAC;AAChD,cAAc,WAAW,CAAC;AAE1B,OAAO,EAAE,QAAQ,EAAE,CAAC;AAEpB,OAAO,EAAE,IAAI,EAAE,sBAAsB,EAAE,MAAM,eAAe,CAAC;AAC7D,YAAY,EAAE,YAAY,EAAE,OAAO,EAAE,MAAM,eAAe,CAAC;AAC3D,OAAO,EAAE,UAAU,EAAE,MAAM,iBAAiB,CAAC;AAC7C,YAAY,EAAE,QAAQ,EAAE,WAAW,EAAE,eAAe,EAAE,MAAM,iBAAiB,CAAC;AAE9E,OAAO,EAAE,6BAA6B,EAAE,MAAM,wBAAwB,CAAC;AACvE,YAAY,EAAE,uBAAuB,EAAE,iBAAiB,EAAE,MAAM,wBAAwB,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,oBAAoB,EAAE,MAAM,uBAAuB,CAAC;AAC7D,OAAO,EAAE,sBAAsB,EAAE,MAAM,4BAA4B,CAAC;AACpE,YAAY,EAAE,QAAQ,EAAE,QAAQ,EAAE,aAAa,EAAE,WAAW,EAAE,MAAM,kBAAkB,CAAC;AACvF,cAAc,kBAAkB,CAAC;AACjC,OAAO,EACN,aAAa,EACb,cAAc,EACd,KAAK,UAAU,EACf,KAAK,YAAY,EACjB,KAAK,YAAY,EACjB,KAAK,QAAQ,EACb,UAAU,EACV,eAAe,EACf,aAAa,EACb,kBAAkB,GAClB,MAAM,gBAAgB,CAAC;AACxB,OAAO,EACN,8BAA8B,EAC9B,sBAAsB,EACtB,2BAA2B,GAC3B,MAAM,sBAAsB,CAAC;AAC9B,YAAY,EAAE,gBAAgB,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAC9E,YAAY,EAAE,SAAS,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AACnE,YAAY,EAAE,UAAU,EAAE,gBAAgB,EAAE,eAAe,EAAE,MAAM,qBAAqB,CAAC;AACzF,YAAY,EACX,aAAa,EACb,sBAAsB,EACtB,gCAAgC,EAChC,8BAA8B,EAC9B,kCAAkC,EAClC,uBAAuB,EACvB,sBAAsB,EACtB,oCAAoC,EACpC,gCAAgC,EAChC,2BAA2B,EAC3B,gCAAgC,EAChC,8BAA8B,EAC9B,sBAAsB,EACtB,8BAA8B,EAC9B,mBAAmB,EACnB,sBAAsB,EACtB,yBAAyB,EACzB,8BAA8B,EAC9B,uBAAuB,GACvB,MAAM,qBAAqB,CAAC;AAC7B,OAAO,EAAE,yBAAyB,EAAE,MAAM,qBAAqB,CAAC;AAEhE,OAAO,KAAK,QAAQ,MAAM,qBAAqB,CAAC;AAChD,cAAc,WAAW,CAAC;AAE1B,OAAO,EAAE,QAAQ,EAAE,CAAC;AACpB,YAAY,EAAE,gBAAgB,EAAE,MAAM,qBAAqB,CAAC;AAG5D,OAAO,EAAE,IAAI,EAAE,sBAAsB,EAAE,MAAM,eAAe,CAAC;AAC7D,YAAY,EAAE,YAAY,EAAE,OAAO,EAAE,MAAM,eAAe,CAAC;AAC3D,OAAO,EAAE,UAAU,EAAE,MAAM,iBAAiB,CAAC;AAC7C,YAAY,EAAE,QAAQ,EAAE,WAAW,EAAE,eAAe,EAAE,MAAM,iBAAiB,CAAC;AAE9E,OAAO,EAAE,6BAA6B,EAAE,MAAM,wBAAwB,CAAC;AACvE,YAAY,EAAE,uBAAuB,EAAE,iBAAiB,EAAE,MAAM,wBAAwB,CAAC"}

package/dist/esm/pipelines.d.ts

@@ -170,6 +170,12 @@ export declare const PIPELINE_DATA: {
         modality: "audio";
         color: "green";
     };
+    "audio-text-to-text": {
+        name: string;
+        modality: "multimodal";
+        color: "red";
+        hideInDatasets: true;
+    };
     "voice-activity-detection": {
         name: string;
         modality: "audio";
@@ -421,5 +427,5 @@ export type PipelineType = keyof typeof PIPELINE_DATA;
 export type WidgetType = PipelineType | "conversational";
 export declare const PIPELINE_TYPES: PipelineType[];
 export declare const SUBTASK_TYPES: string[];
-export declare const PIPELINE_TYPES_SET: Set<"other" | "text-classification" | "token-classification" | "table-question-answering" | "question-answering" | "zero-shot-classification" | "translation" | "summarization" | "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" | "image-text-to-text" | "visual-question-answering" | "document-question-answering" | "zero-shot-image-classification" | "graph-ml" | "mask-generation" | "zero-shot-object-detection" | "text-to-3d" | "image-to-3d" | "image-feature-extraction" | "video-text-to-text" | "keypoint-detection" | "any-to-any">;
+export declare const PIPELINE_TYPES_SET: Set<"other" | "text-classification" | "token-classification" | "table-question-answering" | "question-answering" | "zero-shot-classification" | "translation" | "summarization" | "feature-extraction" | "text-generation" | "text2text-generation" | "fill-mask" | "sentence-similarity" | "text-to-speech" | "text-to-audio" | "automatic-speech-recognition" | "audio-to-audio" | "audio-classification" | "audio-text-to-text" | "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" | "image-text-to-text" | "visual-question-answering" | "document-question-answering" | "zero-shot-image-classification" | "graph-ml" | "mask-generation" | "zero-shot-object-detection" | "text-to-3d" | "image-to-3d" | "image-feature-extraction" | "video-text-to-text" | "keypoint-detection" | "any-to-any">;
 //# sourceMappingURL=pipelines.d.ts.map

package/dist/esm/pipelines.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"pipelines.d.ts","sourceRoot":"","sources":["../../src/pipelines.ts"],"names":[],"mappings":"AAAA,eAAO,MAAM,UAAU,yEAA0E,CAAC;AAElG,MAAM,MAAM,QAAQ,GAAG,CAAC,OAAO,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC;AAEnD,eAAO,MAAM,eAAe;;;;;;;;CAQQ,CAAC;AAErC;;;;;;GAMG;AACH,MAAM,WAAW,OAAO;IACvB;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;IACb;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;CACb;AAED;;;;;GAKG;AACH,MAAM,WAAW,YAAY;IAC5B;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,CAAC,EAAE,OAAO,EAAE,CAAC;IACrB,QAAQ,EAAE,QAAQ,CAAC;IACnB;;OAEG;IACH,KAAK,EAAE,MAAM,GAAG,OAAO,GAAG,QAAQ,GAAG,QAAQ,GAAG,KAAK,GAAG,QAAQ,CAAC;IACjE;;OAEG;IACH,YAAY,CAAC,EAAE,OAAO,CAAC;IACvB;;OAEG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;CACzB;AAcD,eAAO,MAAM,aAAa;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;CAsmBc,CAAC;AAEzC,MAAM,MAAM,YAAY,GAAG,MAAM,OAAO,aAAa,CAAC;AAEtD,MAAM,MAAM,UAAU,GAAG,YAAY,GAAG,gBAAgB,CAAC;AAEzD,eAAO,MAAM,cAAc,EAAiC,YAAY,EAAE,CAAC;AAE3E,eAAO,MAAM,aAAa,UAEN,CAAC;AAErB,eAAO,MAAM,kBAAkB,ipCAA0B,CAAC"}
+{"version":3,"file":"pipelines.d.ts","sourceRoot":"","sources":["../../src/pipelines.ts"],"names":[],"mappings":"AAAA,eAAO,MAAM,UAAU,yEAA0E,CAAC;AAElG,MAAM,MAAM,QAAQ,GAAG,CAAC,OAAO,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC;AAEnD,eAAO,MAAM,eAAe;;;;;;;;CAQQ,CAAC;AAErC;;;;;;GAMG;AACH,MAAM,WAAW,OAAO;IACvB;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;IACb;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;CACb;AAED;;;;;GAKG;AACH,MAAM,WAAW,YAAY;IAC5B;;OAEG;IACH,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,CAAC,EAAE,OAAO,EAAE,CAAC;IACrB,QAAQ,EAAE,QAAQ,CAAC;IACnB;;OAEG;IACH,KAAK,EAAE,MAAM,GAAG,OAAO,GAAG,QAAQ,GAAG,QAAQ,GAAG,KAAK,GAAG,QAAQ,CAAC;IACjE;;OAEG;IACH,YAAY,CAAC,EAAE,OAAO,CAAC;IACvB;;OAEG;IACH,cAAc,CAAC,EAAE,OAAO,CAAC;CACzB;AAcD,eAAO,MAAM,aAAa;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;CA4mBc,CAAC;AAEzC,MAAM,MAAM,YAAY,GAAG,MAAM,OAAO,aAAa,CAAC;AAEtD,MAAM,MAAM,UAAU,GAAG,YAAY,GAAG,gBAAgB,CAAC;AAEzD,eAAO,MAAM,cAAc,EAAiC,YAAY,EAAE,CAAC;AAE3E,eAAO,MAAM,aAAa,UAEN,CAAC;AAErB,eAAO,MAAM,kBAAkB,wqCAA0B,CAAC"}

package/dist/esm/pipelines.js

@@ -306,6 +306,12 @@ export const PIPELINE_DATA = {
         modality: "audio",
         color: "green",
     },
+    "audio-text-to-text": {
+        name: "Audio-Text-to-Text",
+        modality: "multimodal",
+        color: "red",
+        hideInDatasets: true,
+    },
     "voice-activity-detection": {
         name: "Voice Activity Detection",
         modality: "audio",

package/dist/esm/tasks/automatic-speech-recognition/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/automatic-speech-recognition/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cA6Ef,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/automatic-speech-recognition/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAyFf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/esm/tasks/automatic-speech-recognition/data.js

@@ -4,12 +4,16 @@ const taskData = {
             description: "31,175 hours of multilingual audio-text dataset in 108 languages.",
             id: "mozilla-foundation/common_voice_17_0",
         },
+        {
+            description: "Multilingual and diverse audio dataset with 101k hours of audio.",
+            id: "amphion/Emilia-Dataset",
+        },
         {
             description: "A dataset with 44.6k hours of English speaker data and 6k hours of other language speakers.",
             id: "parler-tts/mls_eng",
         },
         {
-            description: "A multi-lingual audio dataset with 370K hours of audio.",
+            description: "A multilingual audio dataset with 370K hours of audio.",
             id: "espnet/yodas",
         },
     ],
@@ -52,6 +56,10 @@ const taskData = {
             description: "An end-to-end model that performs ASR and Speech Translation by MetaAI.",
             id: "facebook/seamless-m4t-v2-large",
         },
+        {
+            description: "A powerful multilingual ASR and Speech Translation model by Nvidia.",
+            id: "nvidia/canary-1b",
+        },
         {
             description: "Powerful speaker diarization model.",
             id: "pyannote/speaker-diarization-3.1",
@@ -63,13 +71,17 @@ const taskData = {
             id: "hf-audio/whisper-large-v3",
         },
         {
-            description: "Fastest speech recognition application.",
-            id: "sanchit-gandhi/whisper-jax",
+            description: "Latest ASR model from Useful Sensors.",
+            id: "mrfakename/Moonshinex",
         },
         {
             description: "A high quality speech and text translation model by Meta.",
             id: "facebook/seamless_m4t",
         },
+        {
+            description: "A powerful multilingual ASR and Speech Translation model by Nvidia",
+            id: "nvidia/canary-1b",
+        },
     ],
     summary: "Automatic Speech Recognition (ASR), also known as Speech to Text (STT), is the task of transcribing a given audio to text. It has many applications, such as voice user interfaces.",
     widgetModels: ["openai/whisper-large-v3"],

package/dist/esm/tasks/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/tasks/index.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAC;AA4CpD,mBAAmB,qCAAqC,CAAC;AACzD,mBAAmB,6CAA6C,CAAC;AACjE,YAAY,EACX,mBAAmB,EACnB,0BAA0B,EAC1B,oBAAoB,EACpB,4BAA4B,EAC5B,2BAA2B,EAC3B,0BAA0B,EAC1B,gCAAgC,EAChC,+BAA+B,GAC/B,MAAM,gCAAgC,CAAC;AACxC,mBAAmB,4CAA4C,CAAC;AAChE,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,0BAA0B,CAAC;AAC9C,YAAY,EACX,wBAAwB,EACxB,yBAAyB,EACzB,gCAAgC,EAChC,6BAA6B,GAC7B,MAAM,qCAAqC,CAAC;AAC7C,mBAAmB,+BAA+B,CAAC;AACnD,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,qBAAqB,EAAE,MAAM,8BAA8B,CAAC;AAC/G,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,iCAAiC,CAAC;AACrD,mBAAmB,iCAAiC,CAAC;AACrD,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,oCAAoC,CAAC;AACxD,mBAAmB,8BAA8B,CAAC;AAClD,mBAAmB,yCAAyC,CAAC;AAC7D,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,qBAAqB,EAAE,MAAM,8BAA8B,CAAC;AAC/G,YAAY,EAAE,sBAAsB,EAAE,iBAAiB,EAAE,kBAAkB,EAAE,MAAM,+BAA+B,CAAC;AACnH,mBAAmB,qCAAqC,CAAC;AACzD,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,MAAM,4BAA4B,CAAC;AACtF,YAAY,EACX,6BAA6B,EAC7B,uBAAuB,EACvB,wBAAwB,EACxB,+BAA+B,EAC/B,4BAA4B,GAC5B,MAAM,oCAAoC,CAAC;AAC5C,YAAY,EACX,gCAAgC,EAChC,gCAAgC,EAChC,mBAAmB,EACnB,oBAAoB,EACpB,2BAA2B,EAC3B,qCAAqC,EACrC,kCAAkC,EAClC,yBAAyB,EACzB,uCAAuC,EACvC,0BAA0B,GAC1B,MAAM,gCAAgC,CAAC;AACxC,mBAAmB,qCAAqC,CAAC;AACzD,mBAAmB,0CAA0C,CAAC;AAC9D,mBAAmB,yCAAyC,CAAC;AAC7D,mBAAmB,+CAA+C,CAAC;AACnE,YAAY,EACX,WAAW,EACX,4BAA4B,EAC5B,gCAAgC,EAChC,6BAA6B,EAC7B,oCAAoC,GACpC,MAAM,2CAA2C,CAAC;AAEnD,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,uBAAuB,CAAC;AAE7D;;GAEG;AACH,eAAO,MAAM,qBAAqB,EAAE,MAAM,CAAC,YAAY,EAAE,eAAe,EAAE,CA6DzE,CAAC;AAoBF,eAAO,MAAM,UAAU,EAAE,MAAM,CAAC,YAAY,EAAE,QAAQ,GAAG,SAAS,CAqDxD,CAAC;AAEX,MAAM,WAAW,WAAW;IAC3B,WAAW,EAAE,MAAM,CAAC;IACpB,EAAE,EAAE,MAAM,CAAC;CACX;AAED,MAAM,MAAM,aAAa,GACtB;IACA,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,OAAO,CAAC;CACb,GACD;IACA,IAAI,EAAE,KAAK,CAAC;QACX,KAAK,EAAE,MAAM,CAAC;QACd,KAAK,EAAE,MAAM,CAAC;KACd,CAAC,CAAC;IACH,IAAI,EAAE,OAAO,CAAC;CACb,GACD;IACA,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,KAAK,CAAC;CACX,GACD;IACA,KAAK,EAAE,MAAM,EAAE,EAAE,CAAC;IAClB,IAAI,EAAE,SAAS,CAAC;CACf,GACD;IACA,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACZ,GACD;IACA,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,KAAK,CAAC;QACb,GAAG,EAAE,MAAM,CAAC;QACZ,KAAK,EAAE,MAAM,CAAC;QACd,IAAI,EAAE,MAAM,CAAC;KACb,CAAC,CAAC;IACH,IAAI,EAAE,kBAAkB,CAAC;CACxB,CAAC;AAEL,MAAM,WAAW,QAAQ;IACxB,MAAM,EAAE,aAAa,EAAE,CAAC;IACxB,OAAO,EAAE,aAAa,EAAE,CAAC;CACzB;AAED,MAAM,WAAW,QAAQ;IACxB,QAAQ,EAAE,WAAW,EAAE,CAAC;IACxB,IAAI,EAAE,QAAQ,CAAC;IACf,EAAE,EAAE,YAAY,CAAC;IACjB,WAAW,CAAC,EAAE,YAAY,CAAC;IAC3B,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,KAAK,EAAE,MAAM,CAAC;IACd,SAAS,EAAE,eAAe,EAAE,CAAC;IAC7B,OAAO,EAAE,WAAW,EAAE,CAAC;IACvB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,OAAO,EAAE,MAAM,CAAC;IAChB,YAAY,EAAE,MAAM,EAAE,CAAC;IACvB,SAAS,CAAC,EAAE,MAAM,CAAC;CACnB;AAED,MAAM,MAAM,cAAc,GAAG,IAAI,CAAC,QAAQ,EAAE,IAAI,GAAG,OAAO,GAAG,WAAW,CAAC,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/tasks/index.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAC;AA4CpD,mBAAmB,qCAAqC,CAAC;AACzD,mBAAmB,6CAA6C,CAAC;AACjE,YAAY,EACX,mBAAmB,EACnB,0BAA0B,EAC1B,oBAAoB,EACpB,4BAA4B,EAC5B,2BAA2B,EAC3B,0BAA0B,EAC1B,gCAAgC,EAChC,+BAA+B,GAC/B,MAAM,gCAAgC,CAAC;AACxC,mBAAmB,4CAA4C,CAAC;AAChE,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,0BAA0B,CAAC;AAC9C,YAAY,EACX,wBAAwB,EACxB,yBAAyB,EACzB,gCAAgC,EAChC,6BAA6B,GAC7B,MAAM,qCAAqC,CAAC;AAC7C,mBAAmB,+BAA+B,CAAC;AACnD,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,qBAAqB,EAAE,MAAM,8BAA8B,CAAC;AAC/G,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,iCAAiC,CAAC;AACrD,mBAAmB,iCAAiC,CAAC;AACrD,mBAAmB,mCAAmC,CAAC;AACvD,mBAAmB,oCAAoC,CAAC;AACxD,mBAAmB,8BAA8B,CAAC;AAClD,mBAAmB,yCAAyC,CAAC;AAC7D,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,qBAAqB,EAAE,MAAM,8BAA8B,CAAC;AAC/G,YAAY,EAAE,sBAAsB,EAAE,iBAAiB,EAAE,kBAAkB,EAAE,MAAM,+BAA+B,CAAC;AACnH,mBAAmB,qCAAqC,CAAC;AACzD,YAAY,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,MAAM,4BAA4B,CAAC;AACtF,YAAY,EACX,6BAA6B,EAC7B,uBAAuB,EACvB,wBAAwB,EACxB,+BAA+B,EAC/B,4BAA4B,GAC5B,MAAM,oCAAoC,CAAC;AAC5C,YAAY,EACX,gCAAgC,EAChC,gCAAgC,EAChC,mBAAmB,EACnB,oBAAoB,EACpB,2BAA2B,EAC3B,qCAAqC,EACrC,kCAAkC,EAClC,yBAAyB,EACzB,uCAAuC,EACvC,0BAA0B,GAC1B,MAAM,gCAAgC,CAAC;AACxC,mBAAmB,qCAAqC,CAAC;AACzD,mBAAmB,0CAA0C,CAAC;AAC9D,mBAAmB,yCAAyC,CAAC;AAC7D,mBAAmB,+CAA+C,CAAC;AACnE,YAAY,EACX,WAAW,EACX,4BAA4B,EAC5B,gCAAgC,EAChC,6BAA6B,EAC7B,oCAAoC,GACpC,MAAM,2CAA2C,CAAC;AAEnD,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,uBAAuB,CAAC;AAE7D;;GAEG;AACH,eAAO,MAAM,qBAAqB,EAAE,MAAM,CAAC,YAAY,EAAE,eAAe,EAAE,CA8DzE,CAAC;AAoBF,eAAO,MAAM,UAAU,EAAE,MAAM,CAAC,YAAY,EAAE,QAAQ,GAAG,SAAS,CAsDxD,CAAC;AAEX,MAAM,WAAW,WAAW;IAC3B,WAAW,EAAE,MAAM,CAAC;IACpB,EAAE,EAAE,MAAM,CAAC;CACX;AAED,MAAM,MAAM,aAAa,GACtB;IACA,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,OAAO,CAAC;CACb,GACD;IACA,IAAI,EAAE,KAAK,CAAC;QACX,KAAK,EAAE,MAAM,CAAC;QACd,KAAK,EAAE,MAAM,CAAC;KACd,CAAC,CAAC;IACH,IAAI,EAAE,OAAO,CAAC;CACb,GACD;IACA,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,KAAK,CAAC;CACX,GACD;IACA,KAAK,EAAE,MAAM,EAAE,EAAE,CAAC;IAClB,IAAI,EAAE,SAAS,CAAC;CACf,GACD;IACA,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACZ,GACD;IACA,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,KAAK,CAAC;QACb,GAAG,EAAE,MAAM,CAAC;QACZ,KAAK,EAAE,MAAM,CAAC;QACd,IAAI,EAAE,MAAM,CAAC;KACb,CAAC,CAAC;IACH,IAAI,EAAE,kBAAkB,CAAC;CACxB,CAAC;AAEL,MAAM,WAAW,QAAQ;IACxB,MAAM,EAAE,aAAa,EAAE,CAAC;IACxB,OAAO,EAAE,aAAa,EAAE,CAAC;CACzB;AAED,MAAM,WAAW,QAAQ;IACxB,QAAQ,EAAE,WAAW,EAAE,CAAC;IACxB,IAAI,EAAE,QAAQ,CAAC;IACf,EAAE,EAAE,YAAY,CAAC;IACjB,WAAW,CAAC,EAAE,YAAY,CAAC;IAC3B,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,KAAK,EAAE,MAAM,CAAC;IACd,SAAS,EAAE,eAAe,EAAE,CAAC;IAC7B,OAAO,EAAE,WAAW,EAAE,CAAC;IACvB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,OAAO,EAAE,MAAM,CAAC;IAChB,YAAY,EAAE,MAAM,EAAE,CAAC;IACvB,SAAS,CAAC,EAAE,MAAM,CAAC;CACnB;AAED,MAAM,MAAM,cAAc,GAAG,IAAI,CAAC,QAAQ,EAAE,IAAI,GAAG,OAAO,GAAG,WAAW,CAAC,CAAC"}

package/dist/esm/tasks/index.js

@@ -46,6 +46,7 @@ export const TASKS_MODEL_LIBRARIES = {
     "audio-classification": ["speechbrain", "transformers", "transformers.js"],
     "audio-to-audio": ["asteroid", "fairseq", "speechbrain"],
     "automatic-speech-recognition": ["espnet", "nemo", "speechbrain", "transformers", "transformers.js"],
+    "audio-text-to-text": [],
     "depth-estimation": ["transformers", "transformers.js"],
     "document-question-answering": ["transformers", "transformers.js"],
     "feature-extraction": ["sentence-transformers", "transformers", "transformers.js"],
@@ -125,6 +126,7 @@ export const TASKS_DATA = {
     "any-to-any": getData("any-to-any", placeholder),
     "audio-classification": getData("audio-classification", audioClassification),
     "audio-to-audio": getData("audio-to-audio", audioToAudio),
+    "audio-text-to-text": getData("audio-text-to-text", placeholder),
     "automatic-speech-recognition": getData("automatic-speech-recognition", automaticSpeechRecognition),
     "depth-estimation": getData("depth-estimation", depthEstimation),
     "document-question-answering": getData("document-question-answering", documentQuestionAnswering),

package/dist/esm/tasks/mask-generation/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/mask-generation/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAkDf,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/mask-generation/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cAgEf,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/esm/tasks/mask-generation/data.js

@@ -1,5 +1,14 @@
 const taskData = {
-    datasets: [],
+    datasets: [
+        {
+            description: "Widely used benchmark dataset for multiple Vision tasks.",
+            id: "merve/coco2017",
+        },
+        {
+            description: "Medical Imaging dataset of the Human Brain for segmentation and mask generating tasks",
+            id: "rocky93/BraTS_segmentation",
+        },
+    ],
     demo: {
         inputs: [
             {
@@ -14,7 +23,12 @@ const taskData = {
             },
         ],
     },
-    metrics: [],
+    metrics: [
+        {
+            description: "IoU is used to measure the overlap between predicted mask and the ground truth mask.",
+            id: "Intersection over Union (IoU)",
+        },
+    ],
     models: [
         {
             description: "Small yet powerful mask generation model.",

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@huggingface/tasks",
   "packageManager": "pnpm@8.10.5",
-  "version": "0.13.2",
+  "version": "0.13.3",
   "description": "List of ML tasks for huggingface.co/tasks",
   "repository": "https://github.com/huggingface/huggingface.js.git",
   "publishConfig": {

package/src/index.ts

@@ -49,6 +49,8 @@ import * as snippets from "./snippets/index.js";
 export * from "./gguf.js";
 
 export { snippets };
+export type { InferenceSnippet } from "./snippets/index.js";
+
 
 export { SKUS, DEFAULT_MEMORY_OPTIONS } from "./hardware.js";
 export type { HardwareSpec, SkuType } from "./hardware.js";

package/src/pipelines.ts

@@ -355,6 +355,12 @@ export const PIPELINE_DATA = {
 		modality: "audio",
 		color: "green",
 	},
+	"audio-text-to-text": {
+		name: "Audio-Text-to-Text",
+		modality: "multimodal",
+		color: "red",
+		hideInDatasets: true,
+	},
 	"voice-activity-detection": {
 		name: "Voice Activity Detection",
 		modality: "audio",

package/src/tasks/automatic-speech-recognition/data.ts

@@ -6,12 +6,16 @@ const taskData: TaskDataCustom = {
 			description: "31,175 hours of multilingual audio-text dataset in 108 languages.",
 			id: "mozilla-foundation/common_voice_17_0",
 		},
+		{
+			description: "Multilingual and diverse audio dataset with 101k hours of audio.",
+			id: "amphion/Emilia-Dataset",
+		},
 		{
 			description: "A dataset with 44.6k hours of English speaker data and 6k hours of other language speakers.",
 			id: "parler-tts/mls_eng",
 		},
 		{
-			description: "A multi-lingual audio dataset with 370K hours of audio.",
+			description: "A multilingual audio dataset with 370K hours of audio.",
 			id: "espnet/yodas",
 		},
 	],
@@ -54,6 +58,10 @@ const taskData: TaskDataCustom = {
 			description: "An end-to-end model that performs ASR and Speech Translation by MetaAI.",
 			id: "facebook/seamless-m4t-v2-large",
 		},
+		{
+			description: "A powerful multilingual ASR and Speech Translation model by Nvidia.",
+			id: "nvidia/canary-1b",
+		},
 		{
 			description: "Powerful speaker diarization model.",
 			id: "pyannote/speaker-diarization-3.1",
@@ -65,13 +73,17 @@ const taskData: TaskDataCustom = {
 			id: "hf-audio/whisper-large-v3",
 		},
 		{
-			description: "Fastest speech recognition application.",
-			id: "sanchit-gandhi/whisper-jax",
+			description: "Latest ASR model from Useful Sensors.",
+			id: "mrfakename/Moonshinex",
 		},
 		{
 			description: "A high quality speech and text translation model by Meta.",
 			id: "facebook/seamless_m4t",
 		},
+		{
+			description: "A powerful multilingual ASR and Speech Translation model by Nvidia",
+			id: "nvidia/canary-1b",
+		},
 	],
 	summary:
 		"Automatic Speech Recognition (ASR), also known as Speech to Text (STT), is the task of transcribing a given audio to text. It has many applications, such as voice user interfaces.",

package/src/tasks/index.ts

@@ -116,6 +116,7 @@ export const TASKS_MODEL_LIBRARIES: Record<PipelineType, ModelLibraryKey[]> = {
 	"audio-classification": ["speechbrain", "transformers", "transformers.js"],
 	"audio-to-audio": ["asteroid", "fairseq", "speechbrain"],
 	"automatic-speech-recognition": ["espnet", "nemo", "speechbrain", "transformers", "transformers.js"],
+	"audio-text-to-text": [],
 	"depth-estimation": ["transformers", "transformers.js"],
 	"document-question-answering": ["transformers", "transformers.js"],
 	"feature-extraction": ["sentence-transformers", "transformers", "transformers.js"],
@@ -197,6 +198,7 @@ export const TASKS_DATA: Record<PipelineType, TaskData | undefined> = {
 	"any-to-any": getData("any-to-any", placeholder),
 	"audio-classification": getData("audio-classification", audioClassification),
 	"audio-to-audio": getData("audio-to-audio", audioToAudio),
+	"audio-text-to-text": getData("audio-text-to-text", placeholder),
 	"automatic-speech-recognition": getData("automatic-speech-recognition", automaticSpeechRecognition),
 	"depth-estimation": getData("depth-estimation", depthEstimation),
 	"document-question-answering": getData("document-question-answering", documentQuestionAnswering),

package/src/tasks/mask-generation/about.md

@@ -12,6 +12,16 @@ Generating masks can facilitate learning, especially in semi or unsupervised lea
 
 For applications where humans are in the loop, masks highlight certain regions of images for humans to validate.
 
+### Medical Imaging
+
+Mask generation models are used in medical imaging to aid in segmenting and analyzing specific regions.
+
+### Autonomous Vehicles
+
+Mask generation models are used to create segments and masks for obstacles and other objects in view.
+
+This page was made possible thanks to the efforts of [Raj Aryan](https://huggingface.co/thatrajaryan) and other contributors.
+
 ## Task Variants
 
 ### Segmentation

package/src/tasks/mask-generation/data.ts

@@ -1,7 +1,16 @@
 import type { TaskDataCustom } from "../index.js";
 
 const taskData: TaskDataCustom = {
-	datasets: [],
+	datasets: [
+		{
+			description: "Widely used benchmark dataset for multiple Vision tasks.",
+			id: "merve/coco2017",
+		},
+		{
+			description: "Medical Imaging dataset of the Human Brain for segmentation and mask generating tasks",
+			id: "rocky93/BraTS_segmentation",
+		},
+	],
 	demo: {
 		inputs: [
 			{
@@ -16,7 +25,12 @@ const taskData: TaskDataCustom = {
 			},
 		],
 	},
-	metrics: [],
+	metrics: [
+		{
+			description: "IoU is used to measure the overlap between predicted mask and the ground truth mask.",
+			id: "Intersection over Union (IoU)",
+		},
+	],
 	models: [
 		{
 			description: "Small yet powerful mask generation model.",