@huggingface/tasks 0.13.0 → 0.13.1

package/dist/index.cjs

@@ -5938,7 +5938,7 @@ var MODEL_LIBRARIES_UI_ELEMENTS = {
     prettyLabel: "open-oasis",
     repoName: "open-oasis",
     repoUrl: "https://github.com/etched-ai/open-oasis",
-    countDownloads: `path:"oasis500m.pt"`
+    countDownloads: `path:"oasis500m.safetensors"`
   },
   open_clip: {
     prettyLabel: "OpenCLIP",
@@ -6524,8 +6524,8 @@ var curlSnippets = {
   "object-detection": snippetFile,
   "image-segmentation": snippetFile
 };
-function getCurlInferenceSnippet(model, accessToken) {
-  return model.pipeline_tag && model.pipeline_tag in curlSnippets ? curlSnippets[model.pipeline_tag]?.(model, accessToken) ?? { content: "" } : { content: "" };
+function getCurlInferenceSnippet(model, accessToken, opts) {
+  return model.pipeline_tag && model.pipeline_tag in curlSnippets ? curlSnippets[model.pipeline_tag]?.(model, accessToken, opts) ?? { content: "" } : { content: "" };
 }
 function hasCurlInferenceSnippet(model) {
   return !!model.pipeline_tag && model.pipeline_tag in curlSnippets;
@@ -6905,7 +6905,7 @@ for await (const chunk of stream) {
       return [
         {
           client: "huggingface.js",
-          content: `import { HfInference } from '@huggingface/inference'
+          content: `import { HfInference } from "@huggingface/inference"
 
 const client = new HfInference("${accessToken || `{API_TOKEN}`}")
 
@@ -7067,8 +7067,8 @@ var jsSnippets = {
   "object-detection": snippetFile3,
   "image-segmentation": snippetFile3
 };
-function getJsInferenceSnippet(model, accessToken) {
-  return model.pipeline_tag && model.pipeline_tag in jsSnippets ? jsSnippets[model.pipeline_tag]?.(model, accessToken) ?? { content: "" } : { content: "" };
+function getJsInferenceSnippet(model, accessToken, opts) {
+  return model.pipeline_tag && model.pipeline_tag in jsSnippets ? jsSnippets[model.pipeline_tag]?.(model, accessToken, opts) ?? { content: "" } : { content: "" };
 }
 function hasJsInferenceSnippet(model) {
   return !!model.pipeline_tag && model.pipeline_tag in jsSnippets;

package/dist/index.js

@@ -5896,7 +5896,7 @@ var MODEL_LIBRARIES_UI_ELEMENTS = {
     prettyLabel: "open-oasis",
     repoName: "open-oasis",
     repoUrl: "https://github.com/etched-ai/open-oasis",
-    countDownloads: `path:"oasis500m.pt"`
+    countDownloads: `path:"oasis500m.safetensors"`
   },
   open_clip: {
     prettyLabel: "OpenCLIP",
@@ -6482,8 +6482,8 @@ var curlSnippets = {
   "object-detection": snippetFile,
   "image-segmentation": snippetFile
 };
-function getCurlInferenceSnippet(model, accessToken) {
-  return model.pipeline_tag && model.pipeline_tag in curlSnippets ? curlSnippets[model.pipeline_tag]?.(model, accessToken) ?? { content: "" } : { content: "" };
+function getCurlInferenceSnippet(model, accessToken, opts) {
+  return model.pipeline_tag && model.pipeline_tag in curlSnippets ? curlSnippets[model.pipeline_tag]?.(model, accessToken, opts) ?? { content: "" } : { content: "" };
 }
 function hasCurlInferenceSnippet(model) {
   return !!model.pipeline_tag && model.pipeline_tag in curlSnippets;
@@ -6863,7 +6863,7 @@ for await (const chunk of stream) {
       return [
         {
           client: "huggingface.js",
-          content: `import { HfInference } from '@huggingface/inference'
+          content: `import { HfInference } from "@huggingface/inference"
 
 const client = new HfInference("${accessToken || `{API_TOKEN}`}")
 
@@ -7025,8 +7025,8 @@ var jsSnippets = {
   "object-detection": snippetFile3,
   "image-segmentation": snippetFile3
 };
-function getJsInferenceSnippet(model, accessToken) {
-  return model.pipeline_tag && model.pipeline_tag in jsSnippets ? jsSnippets[model.pipeline_tag]?.(model, accessToken) ?? { content: "" } : { content: "" };
+function getJsInferenceSnippet(model, accessToken, opts) {
+  return model.pipeline_tag && model.pipeline_tag in jsSnippets ? jsSnippets[model.pipeline_tag]?.(model, accessToken, opts) ?? { content: "" } : { content: "" };
 }
 function hasJsInferenceSnippet(model) {
   return !!model.pipeline_tag && model.pipeline_tag in jsSnippets;

package/dist/src/snippets/curl.d.ts

@@ -12,6 +12,6 @@ export declare const snippetTextGeneration: (model: ModelDataMinimal, accessToke
 export declare const snippetZeroShotClassification: (model: ModelDataMinimal, accessToken: string) => InferenceSnippet;
 export declare const snippetFile: (model: ModelDataMinimal, accessToken: string) => InferenceSnippet;
 export declare const curlSnippets: Partial<Record<PipelineType, (model: ModelDataMinimal, accessToken: string, opts?: Record<string, unknown>) => InferenceSnippet>>;
-export declare function getCurlInferenceSnippet(model: ModelDataMinimal, accessToken: string): InferenceSnippet;
+export declare function getCurlInferenceSnippet(model: ModelDataMinimal, accessToken: string, opts?: Record<string, unknown>): InferenceSnippet;
 export declare function hasCurlInferenceSnippet(model: Pick<ModelDataMinimal, "pipeline_tag">): boolean;
 //# sourceMappingURL=curl.d.ts.map

package/dist/src/snippets/curl.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"curl.d.ts","sourceRoot":"","sources":["../../../src/snippets/curl.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,YAAY,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAM1E,CAAC;AAEH,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,gBAkCF,CAAC;AAEF,eAAO,MAAM,6BAA6B,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAM3F,CAAC;AAEH,eAAO,MAAM,WAAW,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAKzE,CAAC;AAEH,eAAO,MAAM,YAAY,EAAE,OAAO,CACjC,MAAM,CACL,YAAY,EACZ,CAAC,KAAK,EAAE,gBAAgB,EAAE,WAAW,EAAE,MAAM,EAAE,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,gBAAgB,CAClG,CA0BD,CAAC;AAEF,wBAAgB,uBAAuB,CAAC,KAAK,EAAE,gBAAgB,EAAE,WAAW,EAAE,MAAM,GAAG,gBAAgB,CAItG;AAED,wBAAgB,uBAAuB,CAAC,KAAK,EAAE,IAAI,CAAC,gBAAgB,EAAE,cAAc,CAAC,GAAG,OAAO,CAE9F"}
+{"version":3,"file":"curl.d.ts","sourceRoot":"","sources":["../../../src/snippets/curl.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,YAAY,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAM1E,CAAC;AAEH,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,gBAkCF,CAAC;AAEF,eAAO,MAAM,6BAA6B,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAM3F,CAAC;AAEH,eAAO,MAAM,WAAW,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAKzE,CAAC;AAEH,eAAO,MAAM,YAAY,EAAE,OAAO,CACjC,MAAM,CACL,YAAY,EACZ,CAAC,KAAK,EAAE,gBAAgB,EAAE,WAAW,EAAE,MAAM,EAAE,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,gBAAgB,CAClG,CA0BD,CAAC;AAEF,wBAAgB,uBAAuB,CACtC,KAAK,EAAE,gBAAgB,EACvB,WAAW,EAAE,MAAM,EACnB,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC5B,gBAAgB,CAIlB;AAED,wBAAgB,uBAAuB,CAAC,KAAK,EAAE,IAAI,CAAC,gBAAgB,EAAE,cAAc,CAAC,GAAG,OAAO,CAE9F"}

package/dist/src/snippets/js.d.ts

@@ -14,6 +14,6 @@ export declare const snippetTextToImage: (model: ModelDataMinimal, accessToken:
 export declare const snippetTextToAudio: (model: ModelDataMinimal, accessToken: string) => InferenceSnippet;
 export declare const snippetFile: (model: ModelDataMinimal, accessToken: string) => InferenceSnippet;
 export declare const jsSnippets: Partial<Record<PipelineType, (model: ModelDataMinimal, accessToken: string, opts?: Record<string, unknown>) => InferenceSnippet | InferenceSnippet[]>>;
-export declare function getJsInferenceSnippet(model: ModelDataMinimal, accessToken: string): InferenceSnippet | InferenceSnippet[];
+export declare function getJsInferenceSnippet(model: ModelDataMinimal, accessToken: string, opts?: Record<string, unknown>): InferenceSnippet | InferenceSnippet[];
 export declare function hasJsInferenceSnippet(model: ModelDataMinimal): boolean;
 //# sourceMappingURL=js.d.ts.map

package/dist/src/snippets/js.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"js.d.ts","sourceRoot":"","sources":["../../../src/snippets/js.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,YAAY,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAoB1E,CAAC;AAEH,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,GAAG,gBAAgB,EA2GrC,CAAC;AAEF,eAAO,MAAM,6BAA6B,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAsB3F,CAAC;AAEH,eAAO,MAAM,kBAAkB,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAmBhF,CAAC;AAEH,eAAO,MAAM,kBAAkB,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAuCjF,CAAC;AAEF,eAAO,MAAM,WAAW,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAqBzE,CAAC;AAEH,eAAO,MAAM,UAAU,EAAE,OAAO,CAC/B,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,CA0BD,CAAC;AAEF,wBAAgB,qBAAqB,CACpC,KAAK,EAAE,gBAAgB,EACvB,WAAW,EAAE,MAAM,GACjB,gBAAgB,GAAG,gBAAgB,EAAE,CAIvC;AAED,wBAAgB,qBAAqB,CAAC,KAAK,EAAE,gBAAgB,GAAG,OAAO,CAEtE"}
+{"version":3,"file":"js.d.ts","sourceRoot":"","sources":["../../../src/snippets/js.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,YAAY,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAoB1E,CAAC;AAEH,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,GAAG,gBAAgB,EA2GrC,CAAC;AAEF,eAAO,MAAM,6BAA6B,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAsB3F,CAAC;AAEH,eAAO,MAAM,kBAAkB,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAmBhF,CAAC;AAEH,eAAO,MAAM,kBAAkB,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAuCjF,CAAC;AAEF,eAAO,MAAM,WAAW,UAAW,gBAAgB,eAAe,MAAM,KAAG,gBAqBzE,CAAC;AAEH,eAAO,MAAM,UAAU,EAAE,OAAO,CAC/B,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,CA0BD,CAAC;AAEF,wBAAgB,qBAAqB,CACpC,KAAK,EAAE,gBAAgB,EACvB,WAAW,EAAE,MAAM,EACnB,IAAI,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC5B,gBAAgB,GAAG,gBAAgB,EAAE,CAIvC;AAED,wBAAgB,qBAAqB,CAAC,KAAK,EAAE,gBAAgB,GAAG,OAAO,CAEtE"}

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@huggingface/tasks",
   "packageManager": "pnpm@8.10.5",
-  "version": "0.13.0",
+  "version": "0.13.1",
   "description": "List of ML tasks for huggingface.co/tasks",
   "repository": "https://github.com/huggingface/huggingface.js.git",
   "publishConfig": {

package/src/model-libraries.ts

@@ -480,7 +480,7 @@ export const MODEL_LIBRARIES_UI_ELEMENTS = {
 		prettyLabel: "open-oasis",
 		repoName: "open-oasis",
 		repoUrl: "https://github.com/etched-ai/open-oasis",
-		countDownloads: `path:"oasis500m.pt"`,
+		countDownloads: `path:"oasis500m.safetensors"`,
 	},
 	open_clip: {
 		prettyLabel: "OpenCLIP",

package/src/snippets/curl.spec.ts

@@ -1,6 +1,6 @@
 import type { ModelDataMinimal } from "./types";
 import { describe, expect, it } from "vitest";
-import { snippetTextGeneration } from "./curl";
+import { getCurlInferenceSnippet } from "./curl";
 
 describe("inference API snippets", () => {
 	it("conversational llm", async () => {
@@ -10,7 +10,7 @@ describe("inference API snippets", () => {
 			tags: ["conversational"],
 			inference: "",
 		};
-		const snippet = snippetTextGeneration(model, "api_token");
+		const snippet = getCurlInferenceSnippet(model, "api_token");
 
 		expect(snippet.content)
 			.toEqual(`curl 'https://api-inference.huggingface.co/models/meta-llama/Llama-3.1-8B-Instruct/v1/chat/completions' \\
@@ -29,6 +29,32 @@ describe("inference API snippets", () => {
 }'`);
 	});
 
+	it("conversational llm non-streaming", async () => {
+		const model: ModelDataMinimal = {
+			id: "meta-llama/Llama-3.1-8B-Instruct",
+			pipeline_tag: "text-generation",
+			tags: ["conversational"],
+			inference: "",
+		};
+		const snippet = getCurlInferenceSnippet(model, "api_token", { streaming: false });
+
+		expect(snippet.content)
+			.toEqual(`curl 'https://api-inference.huggingface.co/models/meta-llama/Llama-3.1-8B-Instruct/v1/chat/completions' \\
+-H "Authorization: Bearer api_token" \\
+-H 'Content-Type: application/json' \\
+--data '{
+    "model": "meta-llama/Llama-3.1-8B-Instruct",
+    "messages": [
+		{
+			"role": "user",
+			"content": "What is the capital of France?"
+		}
+	],
+    "max_tokens": 500,
+    "stream": false
+}'`);
+	});
+
 	it("conversational vlm", async () => {
 		const model: ModelDataMinimal = {
 			id: "meta-llama/Llama-3.2-11B-Vision-Instruct",
@@ -36,7 +62,7 @@ describe("inference API snippets", () => {
 			tags: ["conversational"],
 			inference: "",
 		};
-		const snippet = snippetTextGeneration(model, "api_token");
+		const snippet = getCurlInferenceSnippet(model, "api_token");
 
 		expect(snippet.content)
 			.toEqual(`curl 'https://api-inference.huggingface.co/models/meta-llama/Llama-3.2-11B-Vision-Instruct/v1/chat/completions' \\

package/src/snippets/curl.ts

@@ -105,9 +105,13 @@ export const curlSnippets: Partial<
 	"image-segmentation": snippetFile,
 };
 
-export function getCurlInferenceSnippet(model: ModelDataMinimal, accessToken: string): InferenceSnippet {
+export function getCurlInferenceSnippet(
+	model: ModelDataMinimal,
+	accessToken: string,
+	opts?: Record<string, unknown>
+): InferenceSnippet {
 	return model.pipeline_tag && model.pipeline_tag in curlSnippets
-		? curlSnippets[model.pipeline_tag]?.(model, accessToken) ?? { content: "" }
+		? curlSnippets[model.pipeline_tag]?.(model, accessToken, opts) ?? { content: "" }
 		: { content: "" };
 }
 

package/src/snippets/js.spec.ts

@@ -1,6 +1,6 @@
 import type { InferenceSnippet, ModelDataMinimal } from "./types";
 import { describe, expect, it } from "vitest";
-import { snippetTextGeneration } from "./js";
+import { getJsInferenceSnippet } from "./js";
 
 describe("inference API snippets", () => {
 	it("conversational llm", async () => {
@@ -10,7 +10,7 @@ describe("inference API snippets", () => {
 			tags: ["conversational"],
 			inference: "",
 		};
-		const snippet = snippetTextGeneration(model, "api_token") as InferenceSnippet[];
+		const snippet = getJsInferenceSnippet(model, "api_token") as InferenceSnippet[];
 
 		expect(snippet[0].content).toEqual(`import { HfInference } from "@huggingface/inference"
 
@@ -38,6 +38,33 @@ for await (const chunk of stream) {
 }`);
 	});
 
+	it("conversational llm non-streaming", async () => {
+		const model: ModelDataMinimal = {
+			id: "meta-llama/Llama-3.1-8B-Instruct",
+			pipeline_tag: "text-generation",
+			tags: ["conversational"],
+			inference: "",
+		};
+		const snippet = getJsInferenceSnippet(model, "api_token", { streaming: false }) as InferenceSnippet[];
+
+		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);`);
+	});
+
 	it("conversational vlm", async () => {
 		const model: ModelDataMinimal = {
 			id: "meta-llama/Llama-3.2-11B-Vision-Instruct",
@@ -45,7 +72,7 @@ for await (const chunk of stream) {
 			tags: ["conversational"],
 			inference: "",
 		};
-		const snippet = snippetTextGeneration(model, "api_token") as InferenceSnippet[];
+		const snippet = getJsInferenceSnippet(model, "api_token") as InferenceSnippet[];
 
 		expect(snippet[0].content).toEqual(`import { HfInference } from "@huggingface/inference"
 
@@ -75,6 +102,41 @@ const stream = client.chatCompletionStream({
 	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);
+	}  
+}`);
+	});
+
+	it("conversational llm", async () => {
+		const model: ModelDataMinimal = {
+			id: "meta-llama/Llama-3.1-8B-Instruct",
+			pipeline_tag: "text-generation",
+			tags: ["conversational"],
+			inference: "",
+		};
+		const snippet = getJsInferenceSnippet(model, "api_token") as InferenceSnippet[];
+
+		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;

package/src/snippets/js.ts

@@ -109,7 +109,7 @@ for await (const chunk of stream) {
 			return [
 				{
 					client: "huggingface.js",
-					content: `import { HfInference } from '@huggingface/inference'
+					content: `import { HfInference } from "@huggingface/inference"
 
 const client = new HfInference("${accessToken || `{API_TOKEN}`}")
 
@@ -292,10 +292,11 @@ export const jsSnippets: Partial<
 
 export function getJsInferenceSnippet(
 	model: ModelDataMinimal,
-	accessToken: string
+	accessToken: string,
+	opts?: Record<string, unknown>
 ): InferenceSnippet | InferenceSnippet[] {
 	return model.pipeline_tag && model.pipeline_tag in jsSnippets
-		? jsSnippets[model.pipeline_tag]?.(model, accessToken) ?? { content: "" }
+		? jsSnippets[model.pipeline_tag]?.(model, accessToken, opts) ?? { content: "" }
 		: { content: "" };
 }
 

package/src/snippets/python.spec.ts

@@ -1,6 +1,6 @@
-import type { ModelDataMinimal } from "./types";
+import type { InferenceSnippet, ModelDataMinimal } from "./types";
 import { describe, expect, it } from "vitest";
-import { snippetConversational } from "./python";
+import { getPythonInferenceSnippet } from "./python";
 
 describe("inference API snippets", () => {
 	it("conversational llm", async () => {
@@ -10,7 +10,7 @@ describe("inference API snippets", () => {
 			tags: ["conversational"],
 			inference: "",
 		};
-		const snippet = snippetConversational(model, "api_token");
+		const snippet = getPythonInferenceSnippet(model, "api_token") as InferenceSnippet[];
 
 		expect(snippet[0].content).toEqual(`from huggingface_hub import InferenceClient
 
@@ -34,6 +34,35 @@ for chunk in stream:
     print(chunk.choices[0].delta.content, end="")`);
 	});
 
+	it("conversational llm non-streaming", async () => {
+		const model: ModelDataMinimal = {
+			id: "meta-llama/Llama-3.1-8B-Instruct",
+			pipeline_tag: "text-generation",
+			tags: ["conversational"],
+			inference: "",
+		};
+		const snippet = getPythonInferenceSnippet(model, "api_token", { streaming: false }) as InferenceSnippet[];
+
+		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)`);
+	});
+
 	it("conversational vlm", async () => {
 		const model: ModelDataMinimal = {
 			id: "meta-llama/Llama-3.2-11B-Vision-Instruct",
@@ -41,7 +70,7 @@ for chunk in stream:
 			tags: ["conversational"],
 			inference: "",
 		};
-		const snippet = snippetConversational(model, "api_token");
+		const snippet = getPythonInferenceSnippet(model, "api_token") as InferenceSnippet[];
 
 		expect(snippet[0].content).toEqual(`from huggingface_hub import InferenceClient