@huggingface/tasks 0.13.1 → 0.13.2

package/dist/commonjs/snippets/js.spec.js

@@ -0,0 +1,141 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+const vitest_1 = require("vitest");
+const js_js_1 = require("./js.js");
+(0, vitest_1.describe)("inference API snippets", () => {
+    (0, vitest_1.it)("conversational llm", async () => {
+        const model = {
+            id: "meta-llama/Llama-3.1-8B-Instruct",
+            pipeline_tag: "text-generation",
+            tags: ["conversational"],
+            inference: "",
+        };
+        const snippet = (0, js_js_1.getJsInferenceSnippet)(model, "api_token");
+        (0, vitest_1.expect)(snippet[0].content).toEqual(`import { HfInference } from "@huggingface/inference";
+
+const client = new HfInference("api_token");
+
+let out = "";
+
+const stream = client.chatCompletionStream({
+	model: "meta-llama/Llama-3.1-8B-Instruct",
+	messages: [
+		{
+			role: "user",
+			content: "What is the capital of France?"
+		}
+	],
+	max_tokens: 500
+});
+
+for await (const chunk of stream) {
+	if (chunk.choices && chunk.choices.length > 0) {
+		const newContent = chunk.choices[0].delta.content;
+		out += newContent;
+		console.log(newContent);
+	}  
+}`);
+    });
+    (0, vitest_1.it)("conversational llm non-streaming", async () => {
+        const model = {
+            id: "meta-llama/Llama-3.1-8B-Instruct",
+            pipeline_tag: "text-generation",
+            tags: ["conversational"],
+            inference: "",
+        };
+        const snippet = (0, js_js_1.getJsInferenceSnippet)(model, "api_token", { streaming: false });
+        (0, vitest_1.expect)(snippet[0].content).toEqual(`import { HfInference } from "@huggingface/inference";
+
+const client = new HfInference("api_token");
+
+const chatCompletion = await client.chatCompletion({
+	model: "meta-llama/Llama-3.1-8B-Instruct",
+	messages: [
+		{
+			role: "user",
+			content: "What is the capital of France?"
+		}
+	],
+	max_tokens: 500
+});
+
+console.log(chatCompletion.choices[0].message);`);
+    });
+    (0, vitest_1.it)("conversational vlm", async () => {
+        const model = {
+            id: "meta-llama/Llama-3.2-11B-Vision-Instruct",
+            pipeline_tag: "image-text-to-text",
+            tags: ["conversational"],
+            inference: "",
+        };
+        const snippet = (0, js_js_1.getJsInferenceSnippet)(model, "api_token");
+        (0, vitest_1.expect)(snippet[0].content).toEqual(`import { HfInference } from "@huggingface/inference";
+
+const client = new HfInference("api_token");
+
+let out = "";
+
+const stream = client.chatCompletionStream({
+	model: "meta-llama/Llama-3.2-11B-Vision-Instruct",
+	messages: [
+		{
+			role: "user",
+			content: [
+				{
+					type: "text",
+					text: "Describe this image in one sentence."
+				},
+				{
+					type: "image_url",
+					image_url: {
+						url: "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
+					}
+				}
+			]
+		}
+	],
+	max_tokens: 500
+});
+
+for await (const chunk of stream) {
+	if (chunk.choices && chunk.choices.length > 0) {
+		const newContent = chunk.choices[0].delta.content;
+		out += newContent;
+		console.log(newContent);
+	}  
+}`);
+    });
+    (0, vitest_1.it)("conversational llm", async () => {
+        const model = {
+            id: "meta-llama/Llama-3.1-8B-Instruct",
+            pipeline_tag: "text-generation",
+            tags: ["conversational"],
+            inference: "",
+        };
+        const snippet = (0, js_js_1.getJsInferenceSnippet)(model, "api_token");
+        (0, vitest_1.expect)(snippet[0].content).toEqual(`import { HfInference } from "@huggingface/inference";
+
+const client = new HfInference("api_token");
+
+let out = "";
+
+const stream = client.chatCompletionStream({
+	model: "meta-llama/Llama-3.1-8B-Instruct",
+	messages: [
+		{
+			role: "user",
+			content: "What is the capital of France?"
+		}
+	],
+	max_tokens: 500
+});
+
+for await (const chunk of stream) {
+	if (chunk.choices && chunk.choices.length > 0) {
+		const newContent = chunk.choices[0].delta.content;
+		out += newContent;
+		console.log(newContent);
+	}  
+}`);
+    });
+});

package/dist/{src → commonjs}/snippets/python.d.ts

@@ -12,7 +12,7 @@ export declare const snippetZeroShotClassification: (model: ModelDataMinimal) =>
 export declare const snippetZeroShotImageClassification: (model: ModelDataMinimal) => InferenceSnippet;
 export declare const snippetBasic: (model: ModelDataMinimal) => InferenceSnippet;
 export declare const snippetFile: (model: ModelDataMinimal) => InferenceSnippet;
-export declare const snippetTextToImage: (model: ModelDataMinimal) => InferenceSnippet;
+export declare const snippetTextToImage: (model: ModelDataMinimal, accessToken: string) => InferenceSnippet[];
 export declare const snippetTabular: (model: ModelDataMinimal) => InferenceSnippet;
 export declare const snippetTextToAudio: (model: ModelDataMinimal) => InferenceSnippet;
 export declare const snippetDocumentQuestionAnswering: (model: ModelDataMinimal) => InferenceSnippet;

package/dist/{src → commonjs}/snippets/python.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"python.d.ts","sourceRoot":"","sources":["../../../src/snippets/python.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAC;AACpD,OAAO,KAAK,EAAE,0BAA0B,EAAE,oBAAoB,EAAE,MAAM,mBAAmB,CAAC;AAG1F,OAAO,KAAK,EAAE,gBAAgB,EAAE,gBAAgB,EAAE,MAAM,YAAY,CAAC;AAErE,eAAO,MAAM,qBAAqB,UAC1B,gBAAgB,eACV,MAAM,SACZ;IACN,SAAS,CAAC,EAAE,OAAO,CAAC;IACpB,QAAQ,CAAC,EAAE,0BAA0B,EAAE,CAAC;IACxC,WAAW,CAAC,EAAE,oBAAoB,CAAC,aAAa,CAAC,CAAC;IAClD,UAAU,CAAC,EAAE,oBAAoB,CAAC,YAAY,CAAC,CAAC;IAChD,KAAK,CAAC,EAAE,oBAAoB,CAAC,OAAO,CAAC,CAAC;CACtC,KACC,gBAAgB,EAiGlB,CAAC;AAEF,eAAO,MAAM,6BAA6B,UAAW,gBAAgB,KAAG,gBAStE,CAAC;AAEH,eAAO,MAAM,kCAAkC,UAAW,gBAAgB,KAAG,gBAe3E,CAAC;AAEH,eAAO,MAAM,YAAY,UAAW,gBAAgB,KAAG,gBAQrD,CAAC;AAEH,eAAO,MAAM,WAAW,UAAW,gBAAgB,KAAG,gBAQpD,CAAC;AAEH,eAAO,MAAM,kBAAkB,UAAW,gBAAgB,KAAG,gBAW3D,CAAC;AAEH,eAAO,MAAM,cAAc,UAAW,gBAAgB,KAAG,gBAOvD,CAAC;AAEH,eAAO,MAAM,kBAAkB,UAAW,gBAAgB,KAAG,gBA+B5D,CAAC;AAEF,eAAO,MAAM,gCAAgC,UAAW,gBAAgB,KAAG,gBAWzE,CAAC;AAEH,eAAO,MAAM,cAAc,EAAE,OAAO,CACnC,MAAM,CACL,YAAY,EACZ,CACC,KAAK,EAAE,gBAAgB,EACvB,WAAW,EAAE,MAAM,EACnB,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,KAC1B,gBAAgB,GAAG,gBAAgB,EAAE,CAC1C,CA8BD,CAAC;AAEF,wBAAgB,yBAAyB,CACxC,KAAK,EAAE,gBAAgB,EACvB,WAAW,EAAE,MAAM,EACnB,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC5B,gBAAgB,GAAG,gBAAgB,EAAE,CAwBvC;AAED,wBAAgB,yBAAyB,CAAC,KAAK,EAAE,gBAAgB,GAAG,OAAO,CAE1E"}
+{"version":3,"file":"python.d.ts","sourceRoot":"","sources":["../../../src/snippets/python.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAC;AACpD,OAAO,KAAK,EAAE,0BAA0B,EAAE,oBAAoB,EAAE,MAAM,mBAAmB,CAAC;AAG1F,OAAO,KAAK,EAAE,gBAAgB,EAAE,gBAAgB,EAAE,MAAM,YAAY,CAAC;AAOrE,eAAO,MAAM,qBAAqB,UAC1B,gBAAgB,eACV,MAAM,SACZ;IACN,SAAS,CAAC,EAAE,OAAO,CAAC;IACpB,QAAQ,CAAC,EAAE,0BAA0B,EAAE,CAAC;IACxC,WAAW,CAAC,EAAE,oBAAoB,CAAC,aAAa,CAAC,CAAC;IAClD,UAAU,CAAC,EAAE,oBAAoB,CAAC,YAAY,CAAC,CAAC;IAChD,KAAK,CAAC,EAAE,oBAAoB,CAAC,OAAO,CAAC,CAAC;CACtC,KACC,gBAAgB,EAiGlB,CAAC;AAEF,eAAO,MAAM,6BAA6B,UAAW,gBAAgB,KAAG,gBAStE,CAAC;AAEH,eAAO,MAAM,kCAAkC,UAAW,gBAAgB,KAAG,gBAe3E,CAAC;AAEH,eAAO,MAAM,YAAY,UAAW,gBAAgB,KAAG,gBAQrD,CAAC;AAEH,eAAO,MAAM,WAAW,UAAW,gBAAgB,KAAG,gBAQpD,CAAC;AAEH,eAAO,MAAM,kBAAkB,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAAgB,EAqBjG,CAAC;AAEF,eAAO,MAAM,cAAc,UAAW,gBAAgB,KAAG,gBAOvD,CAAC;AAEH,eAAO,MAAM,kBAAkB,UAAW,gBAAgB,KAAG,gBA+B5D,CAAC;AAEF,eAAO,MAAM,gCAAgC,UAAW,gBAAgB,KAAG,gBAWzE,CAAC;AAEH,eAAO,MAAM,cAAc,EAAE,OAAO,CACnC,MAAM,CACL,YAAY,EACZ,CACC,KAAK,EAAE,gBAAgB,EACvB,WAAW,EAAE,MAAM,EACnB,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,KAC1B,gBAAgB,GAAG,gBAAgB,EAAE,CAC1C,CA8BD,CAAC;AAEF,wBAAgB,yBAAyB,CACxC,KAAK,EAAE,gBAAgB,EACvB,WAAW,EAAE,MAAM,EACnB,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC5B,gBAAgB,GAAG,gBAAgB,EAAE,CA0BvC;AAED,wBAAgB,yBAAyB,CAAC,KAAK,EAAE,gBAAgB,GAAG,OAAO,CAE1E"}

package/dist/commonjs/snippets/python.js

@@ -0,0 +1,293 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.pythonSnippets = exports.snippetDocumentQuestionAnswering = exports.snippetTextToAudio = exports.snippetTabular = exports.snippetTextToImage = exports.snippetFile = exports.snippetBasic = exports.snippetZeroShotImageClassification = exports.snippetZeroShotClassification = exports.snippetConversational = void 0;
+exports.getPythonInferenceSnippet = getPythonInferenceSnippet;
+exports.hasPythonInferenceSnippet = hasPythonInferenceSnippet;
+const common_js_1 = require("./common.js");
+const inputs_js_1 = require("./inputs.js");
+const snippetImportInferenceClient = (model, accessToken) => `from huggingface_hub import InferenceClient
+client = InferenceClient("${model.id}", token="${accessToken || "{API_TOKEN}"}")
+`;
+const snippetConversational = (model, accessToken, opts) => {
+    const streaming = opts?.streaming ?? true;
+    const exampleMessages = (0, inputs_js_1.getModelInputSnippet)(model);
+    const messages = opts?.messages ?? exampleMessages;
+    const messagesStr = (0, common_js_1.stringifyMessages)(messages, { attributeKeyQuotes: true });
+    const config = {
+        ...(opts?.temperature ? { temperature: opts.temperature } : undefined),
+        max_tokens: opts?.max_tokens ?? 500,
+        ...(opts?.top_p ? { top_p: opts.top_p } : undefined),
+    };
+    const configStr = (0, common_js_1.stringifyGenerationConfig)(config, {
+        indent: "\n\t",
+        attributeValueConnector: "=",
+    });
+    if (streaming) {
+        return [
+            {
+                client: "huggingface_hub",
+                content: `from huggingface_hub import InferenceClient
+
+client = InferenceClient(api_key="${accessToken || "{API_TOKEN}"}")
+
+messages = ${messagesStr}
+
+stream = client.chat.completions.create(
+    model="${model.id}", 
+	messages=messages, 
+	${configStr},
+	stream=True
+)
+
+for chunk in stream:
+    print(chunk.choices[0].delta.content, end="")`,
+            },
+            {
+                client: "openai",
+                content: `from openai import OpenAI
+
+client = OpenAI(
+	base_url="https://api-inference.huggingface.co/v1/",
+	api_key="${accessToken || "{API_TOKEN}"}"
+)
+
+messages = ${messagesStr}
+
+stream = client.chat.completions.create(
+    model="${model.id}", 
+	messages=messages, 
+	${configStr},
+	stream=True
+)
+
+for chunk in stream:
+    print(chunk.choices[0].delta.content, end="")`,
+            },
+        ];
+    }
+    else {
+        return [
+            {
+                client: "huggingface_hub",
+                content: `from huggingface_hub import InferenceClient
+
+client = InferenceClient(api_key="${accessToken || "{API_TOKEN}"}")
+
+messages = ${messagesStr}
+
+completion = client.chat.completions.create(
+    model="${model.id}", 
+	messages=messages, 
+	${configStr}
+)
+
+print(completion.choices[0].message)`,
+            },
+            {
+                client: "openai",
+                content: `from openai import OpenAI
+
+client = OpenAI(
+	base_url="https://api-inference.huggingface.co/v1/",
+	api_key="${accessToken || "{API_TOKEN}"}"
+)
+
+messages = ${messagesStr}
+
+completion = client.chat.completions.create(
+    model="${model.id}", 
+	messages=messages, 
+	${configStr}
+)
+
+print(completion.choices[0].message)`,
+            },
+        ];
+    }
+};
+exports.snippetConversational = snippetConversational;
+const snippetZeroShotClassification = (model) => ({
+    content: `def query(payload):
+	response = requests.post(API_URL, headers=headers, json=payload)
+	return response.json()
+
+output = query({
+    "inputs": ${(0, inputs_js_1.getModelInputSnippet)(model)},
+    "parameters": {"candidate_labels": ["refund", "legal", "faq"]},
+})`,
+});
+exports.snippetZeroShotClassification = snippetZeroShotClassification;
+const snippetZeroShotImageClassification = (model) => ({
+    content: `def query(data):
+	with open(data["image_path"], "rb") as f:
+		img = f.read()
+	payload={
+		"parameters": data["parameters"],
+		"inputs": base64.b64encode(img).decode("utf-8")
+	}
+	response = requests.post(API_URL, headers=headers, json=payload)
+	return response.json()
+
+output = query({
+    "image_path": ${(0, inputs_js_1.getModelInputSnippet)(model)},
+    "parameters": {"candidate_labels": ["cat", "dog", "llama"]},
+})`,
+});
+exports.snippetZeroShotImageClassification = snippetZeroShotImageClassification;
+const snippetBasic = (model) => ({
+    content: `def query(payload):
+	response = requests.post(API_URL, headers=headers, json=payload)
+	return response.json()
+	
+output = query({
+	"inputs": ${(0, inputs_js_1.getModelInputSnippet)(model)},
+})`,
+});
+exports.snippetBasic = snippetBasic;
+const snippetFile = (model) => ({
+    content: `def query(filename):
+    with open(filename, "rb") as f:
+        data = f.read()
+    response = requests.post(API_URL, headers=headers, data=data)
+    return response.json()
+
+output = query(${(0, inputs_js_1.getModelInputSnippet)(model)})`,
+});
+exports.snippetFile = snippetFile;
+const snippetTextToImage = (model, accessToken) => [
+    {
+        client: "huggingface_hub",
+        content: `${snippetImportInferenceClient(model, accessToken)}
+# output is a PIL.Image object
+image = client.text_to_image(${(0, inputs_js_1.getModelInputSnippet)(model)})`,
+    },
+    {
+        client: "requests",
+        content: `def query(payload):
+	response = requests.post(API_URL, headers=headers, json=payload)
+	return response.content
+image_bytes = query({
+	"inputs": ${(0, inputs_js_1.getModelInputSnippet)(model)},
+})
+
+# You can access the image with PIL.Image for example
+import io
+from PIL import Image
+image = Image.open(io.BytesIO(image_bytes))`,
+    },
+];
+exports.snippetTextToImage = snippetTextToImage;
+const snippetTabular = (model) => ({
+    content: `def query(payload):
+	response = requests.post(API_URL, headers=headers, json=payload)
+	return response.content
+response = query({
+	"inputs": {"data": ${(0, inputs_js_1.getModelInputSnippet)(model)}},
+})`,
+});
+exports.snippetTabular = snippetTabular;
+const snippetTextToAudio = (model) => {
+    // Transformers TTS pipeline and api-inference-community (AIC) pipeline outputs are diverged
+    // with the latest update to inference-api (IA).
+    // Transformers IA returns a byte object (wav file), whereas AIC returns wav and sampling_rate.
+    if (model.library_name === "transformers") {
+        return {
+            content: `def query(payload):
+	response = requests.post(API_URL, headers=headers, json=payload)
+	return response.content
+
+audio_bytes = query({
+	"inputs": ${(0, inputs_js_1.getModelInputSnippet)(model)},
+})
+# You can access the audio with IPython.display for example
+from IPython.display import Audio
+Audio(audio_bytes)`,
+        };
+    }
+    else {
+        return {
+            content: `def query(payload):
+	response = requests.post(API_URL, headers=headers, json=payload)
+	return response.json()
+	
+audio, sampling_rate = query({
+	"inputs": ${(0, inputs_js_1.getModelInputSnippet)(model)},
+})
+# You can access the audio with IPython.display for example
+from IPython.display import Audio
+Audio(audio, rate=sampling_rate)`,
+        };
+    }
+};
+exports.snippetTextToAudio = snippetTextToAudio;
+const snippetDocumentQuestionAnswering = (model) => ({
+    content: `def query(payload):
+ 	with open(payload["image"], "rb") as f:
+  		img = f.read()
+		payload["image"] = base64.b64encode(img).decode("utf-8")  
+	response = requests.post(API_URL, headers=headers, json=payload)
+	return response.json()
+
+output = query({
+    "inputs": ${(0, inputs_js_1.getModelInputSnippet)(model)},
+})`,
+});
+exports.snippetDocumentQuestionAnswering = snippetDocumentQuestionAnswering;
+exports.pythonSnippets = {
+    // Same order as in tasks/src/pipelines.ts
+    "text-classification": exports.snippetBasic,
+    "token-classification": exports.snippetBasic,
+    "table-question-answering": exports.snippetBasic,
+    "question-answering": exports.snippetBasic,
+    "zero-shot-classification": exports.snippetZeroShotClassification,
+    translation: exports.snippetBasic,
+    summarization: exports.snippetBasic,
+    "feature-extraction": exports.snippetBasic,
+    "text-generation": exports.snippetBasic,
+    "text2text-generation": exports.snippetBasic,
+    "image-text-to-text": exports.snippetConversational,
+    "fill-mask": exports.snippetBasic,
+    "sentence-similarity": exports.snippetBasic,
+    "automatic-speech-recognition": exports.snippetFile,
+    "text-to-image": exports.snippetTextToImage,
+    "text-to-speech": exports.snippetTextToAudio,
+    "text-to-audio": exports.snippetTextToAudio,
+    "audio-to-audio": exports.snippetFile,
+    "audio-classification": exports.snippetFile,
+    "image-classification": exports.snippetFile,
+    "tabular-regression": exports.snippetTabular,
+    "tabular-classification": exports.snippetTabular,
+    "object-detection": exports.snippetFile,
+    "image-segmentation": exports.snippetFile,
+    "document-question-answering": exports.snippetDocumentQuestionAnswering,
+    "image-to-text": exports.snippetFile,
+    "zero-shot-image-classification": exports.snippetZeroShotImageClassification,
+};
+function getPythonInferenceSnippet(model, accessToken, opts) {
+    if (model.tags.includes("conversational")) {
+        // Conversational model detected, so we display a code snippet that features the Messages API
+        return (0, exports.snippetConversational)(model, accessToken, opts);
+    }
+    else {
+        let snippets = model.pipeline_tag && model.pipeline_tag in exports.pythonSnippets
+            ? exports.pythonSnippets[model.pipeline_tag]?.(model, accessToken) ?? { content: "" }
+            : { content: "" };
+        snippets = Array.isArray(snippets) ? snippets : [snippets];
+        return snippets.map((snippet) => {
+            return {
+                ...snippet,
+                content: snippet.content.includes("requests")
+                    ? `import requests
+
+API_URL = "https://api-inference.huggingface.co/models/${model.id}"
+headers = {"Authorization": ${accessToken ? `"Bearer ${accessToken}"` : `f"Bearer {API_TOKEN}"`}}
+
+${snippet.content}`
+                    : snippet.content,
+            };
+        });
+    }
+}
+function hasPythonInferenceSnippet(model) {
+    return !!model.pipeline_tag && model.pipeline_tag in exports.pythonSnippets;
+}

package/dist/commonjs/snippets/python.spec.js

@@ -0,0 +1,135 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+const vitest_1 = require("vitest");
+const python_js_1 = require("./python.js");
+(0, vitest_1.describe)("inference API snippets", () => {
+    (0, vitest_1.it)("conversational llm", async () => {
+        const model = {
+            id: "meta-llama/Llama-3.1-8B-Instruct",
+            pipeline_tag: "text-generation",
+            tags: ["conversational"],
+            inference: "",
+        };
+        const snippet = (0, python_js_1.getPythonInferenceSnippet)(model, "api_token");
+        (0, vitest_1.expect)(snippet[0].content).toEqual(`from huggingface_hub import InferenceClient
+
+client = InferenceClient(api_key="api_token")
+
+messages = [
+	{
+		"role": "user",
+		"content": "What is the capital of France?"
+	}
+]
+
+stream = client.chat.completions.create(
+    model="meta-llama/Llama-3.1-8B-Instruct", 
+	messages=messages, 
+	max_tokens=500,
+	stream=True
+)
+
+for chunk in stream:
+    print(chunk.choices[0].delta.content, end="")`);
+    });
+    (0, vitest_1.it)("conversational llm non-streaming", async () => {
+        const model = {
+            id: "meta-llama/Llama-3.1-8B-Instruct",
+            pipeline_tag: "text-generation",
+            tags: ["conversational"],
+            inference: "",
+        };
+        const snippet = (0, python_js_1.getPythonInferenceSnippet)(model, "api_token", { streaming: false });
+        (0, vitest_1.expect)(snippet[0].content).toEqual(`from huggingface_hub import InferenceClient
+
+client = InferenceClient(api_key="api_token")
+
+messages = [
+	{
+		"role": "user",
+		"content": "What is the capital of France?"
+	}
+]
+
+completion = client.chat.completions.create(
+    model="meta-llama/Llama-3.1-8B-Instruct", 
+	messages=messages, 
+	max_tokens=500
+)
+
+print(completion.choices[0].message)`);
+    });
+    (0, vitest_1.it)("conversational vlm", async () => {
+        const model = {
+            id: "meta-llama/Llama-3.2-11B-Vision-Instruct",
+            pipeline_tag: "image-text-to-text",
+            tags: ["conversational"],
+            inference: "",
+        };
+        const snippet = (0, python_js_1.getPythonInferenceSnippet)(model, "api_token");
+        (0, vitest_1.expect)(snippet[0].content).toEqual(`from huggingface_hub import InferenceClient
+
+client = InferenceClient(api_key="api_token")
+
+messages = [
+	{
+		"role": "user",
+		"content": [
+			{
+				"type": "text",
+				"text": "Describe this image in one sentence."
+			},
+			{
+				"type": "image_url",
+				"image_url": {
+					"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
+				}
+			}
+		]
+	}
+]
+
+stream = client.chat.completions.create(
+    model="meta-llama/Llama-3.2-11B-Vision-Instruct", 
+	messages=messages, 
+	max_tokens=500,
+	stream=True
+)
+
+for chunk in stream:
+    print(chunk.choices[0].delta.content, end="")`);
+    });
+    (0, vitest_1.it)("text-to-image", async () => {
+        const model = {
+            id: "black-forest-labs/FLUX.1-schnell",
+            pipeline_tag: "text-to-image",
+            tags: [],
+            inference: "",
+        };
+        const snippets = (0, python_js_1.getPythonInferenceSnippet)(model, "api_token");
+        (0, vitest_1.expect)(snippets.length).toEqual(2);
+        (0, vitest_1.expect)(snippets[0].client).toEqual("huggingface_hub");
+        (0, vitest_1.expect)(snippets[0].content).toEqual(`from huggingface_hub import InferenceClient
+client = InferenceClient("black-forest-labs/FLUX.1-schnell", token="api_token")
+
+# output is a PIL.Image object
+image = client.text_to_image("Astronaut riding a horse")`);
+        (0, vitest_1.expect)(snippets[1].client).toEqual("requests");
+        (0, vitest_1.expect)(snippets[1].content).toEqual(`import requests
+
+API_URL = "https://api-inference.huggingface.co/models/black-forest-labs/FLUX.1-schnell"
+headers = {"Authorization": "Bearer api_token"}
+
+def query(payload):
+	response = requests.post(API_URL, headers=headers, json=payload)
+	return response.content
+image_bytes = query({
+	"inputs": "Astronaut riding a horse",
+})
+
+# You can access the image with PIL.Image for example
+import io
+from PIL import Image
+image = Image.open(io.BytesIO(image_bytes))`);
+    });
+});

package/dist/{src → commonjs}/snippets/types.d.ts

@@ -1,4 +1,4 @@
-import type { ModelData } from "../model-data";
+import type { ModelData } from "../model-data.js";
 /**
  * Minimal model data required for snippets.
  *

package/dist/commonjs/snippets/types.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../../../src/snippets/types.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAC;AAElD;;;;GAIG;AACH,MAAM,MAAM,gBAAgB,GAAG,IAAI,CAClC,SAAS,EACT,IAAI,GAAG,cAAc,GAAG,YAAY,GAAG,cAAc,GAAG,QAAQ,GAAG,MAAM,GAAG,WAAW,CACvF,CAAC;AAEF,MAAM,WAAW,gBAAgB;IAChC,OAAO,EAAE,MAAM,CAAC;IAChB,MAAM,CAAC,EAAE,MAAM,CAAC;CAChB"}

package/dist/commonjs/snippets/types.js

@@ -0,0 +1,2 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });

package/dist/commonjs/tasks/audio-classification/data.d.ts

@@ -0,0 +1,4 @@
+import type { TaskDataCustom } from "../index.js";
+declare const taskData: TaskDataCustom;
+export default taskData;
+//# sourceMappingURL=data.d.ts.map

package/dist/{src → commonjs}/tasks/audio-classification/data.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/audio-classification/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,IAAI,CAAC;AAEzC,QAAA,MAAM,QAAQ,EAAE,cA4Ef,CAAC;AAEF,eAAe,QAAQ,CAAC"}
+{"version":3,"file":"data.d.ts","sourceRoot":"","sources":["../../../../src/tasks/audio-classification/data.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAElD,QAAA,MAAM,QAAQ,EAAE,cA4Ef,CAAC;AAEF,eAAe,QAAQ,CAAC"}

package/dist/commonjs/tasks/audio-classification/data.js

@@ -0,0 +1,79 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+const taskData = {
+    datasets: [
+        {
+            description: "A benchmark of 10 different audio tasks.",
+            id: "s3prl/superb",
+        },
+        {
+            description: "A dataset of YouTube clips and their sound categories.",
+            id: "agkphysics/AudioSet",
+        },
+    ],
+    demo: {
+        inputs: [
+            {
+                filename: "audio.wav",
+                type: "audio",
+            },
+        ],
+        outputs: [
+            {
+                data: [
+                    {
+                        label: "Up",
+                        score: 0.2,
+                    },
+                    {
+                        label: "Down",
+                        score: 0.8,
+                    },
+                ],
+                type: "chart",
+            },
+        ],
+    },
+    metrics: [
+        {
+            description: "",
+            id: "accuracy",
+        },
+        {
+            description: "",
+            id: "recall",
+        },
+        {
+            description: "",
+            id: "precision",
+        },
+        {
+            description: "",
+            id: "f1",
+        },
+    ],
+    models: [
+        {
+            description: "An easy-to-use model for command recognition.",
+            id: "speechbrain/google_speech_command_xvector",
+        },
+        {
+            description: "An emotion recognition model.",
+            id: "ehcalabres/wav2vec2-lg-xlsr-en-speech-emotion-recognition",
+        },
+        {
+            description: "A language identification model.",
+            id: "facebook/mms-lid-126",
+        },
+    ],
+    spaces: [
+        {
+            description: "An application that can classify music into different genre.",
+            id: "kurianbenoy/audioclassification",
+        },
+    ],
+    summary: "Audio classification is the task of assigning a label or class to a given audio. It can be used for recognizing which command a user is giving or the emotion of a statement, as well as identifying a speaker.",
+    widgetModels: ["MIT/ast-finetuned-audioset-10-10-0.4593"],
+    youtubeId: "KWwzcmG98Ds",
+};
+exports.default = taskData;