@huggingface/inference 2.6.7 → 2.7.1

package/src/tasks/nlp/textGenerationStream.ts

@@ -1,6 +1,6 @@
+import type { TextGenerationInput } from "@huggingface/tasks";
 import type { BaseArgs, Options } from "../../types";
 import { streamingRequest } from "../custom/streamingRequest";
-import type { TextGenerationInput } from "./textGeneration";
 
 export interface TextGenerationStreamToken {
 	/** Token ID from the model tokenizer */
@@ -67,6 +67,7 @@ export interface TextGenerationStreamDetails {
 }
 
 export interface TextGenerationStreamOutput {
+	index?: number;
 	/** Generated token, one at a time */
 	token: TextGenerationStreamToken;
 	/**

package/src/types.ts

@@ -1,4 +1,5 @@
 import type { PipelineType } from "@huggingface/tasks";
+import type { ChatCompletionInput } from "@huggingface/tasks";
 
 export interface Options {
 	/**
@@ -32,7 +33,7 @@ export interface Options {
 	signal?: AbortSignal;
 
 	/**
-	 * Credentials to use for the request. If this is a string, it will be passed straight on. If it's a boolean, true will be "include" and false will not send credentials at all (which defaults to "same-origin" inside browsers).
+	 * (Default: "same-origin"). String | Boolean. Credentials to use for the request. If this is a string, it will be passed straight on. If it's a boolean, true will be "include" and false will not send credentials at all.
 	 */
 	includeCredentials?: string | boolean;
 }
@@ -47,15 +48,25 @@ export interface BaseArgs {
 	 */
 	accessToken?: string;
 	/**
-	 * The model to use. Can be a full URL for a dedicated inference endpoint.
+	 * The model to use.
 	 *
 	 * If not specified, will call huggingface.co/api/tasks to get the default model for the task.
+	 *
+	 * /!\ Legacy behavior allows this to be an URL, but this is deprecated and will be removed in the future.
+	 * Use the `endpointUrl` parameter instead.
 	 */
 	model?: string;
+
+	/**
+	 * The URL of the endpoint to use. If not specified, will call huggingface.co/api/tasks to get the default endpoint for the task.
+	 *
+	 * If specified, will use this URL instead of the default one.
+	 */
+	endpointUrl?: string;
 }
 
 export type RequestArgs = BaseArgs &
-	({ data: Blob | ArrayBuffer } | { inputs: unknown }) & {
+	({ data: Blob | ArrayBuffer } | { inputs: unknown } | ChatCompletionInput) & {
 		parameters?: Record<string, unknown>;
 		accessToken?: string;
 	};

package/src/utils/base64FromBytes.ts

@@ -0,0 +1,11 @@
+export function base64FromBytes(arr: Uint8Array): string {
+	if (globalThis.Buffer) {
+		return globalThis.Buffer.from(arr).toString("base64");
+	} else {
+		const bin: string[] = [];
+		arr.forEach((byte) => {
+			bin.push(String.fromCharCode(byte));
+		});
+		return globalThis.btoa(bin.join(""));
+	}
+}

package/src/utils/{distributive-omit.d.ts → distributive-omit.ts}

@@ -4,8 +4,6 @@
  * This allows omitting keys from objects inside unions, without merging the individual components of the union.
  */
 
-type Keys<T> = keyof T;
-type DistributiveKeys<T> = T extends unknown ? Keys<T> : never;
 type Omit_<T, K> = Omit<T, Extract<keyof T, K>>;
 
 export type DistributiveOmit<T, K> = T extends unknown

package/src/utils/isBackend.ts

@@ -0,0 +1,6 @@
+const isBrowser = typeof window !== "undefined" && typeof window.document !== "undefined";
+
+const isWebWorker =
+	typeof self === "object" && self.constructor && self.constructor.name === "DedicatedWorkerGlobalScope";
+
+export const isBackend = !isBrowser && !isWebWorker;

package/src/utils/isFrontend.ts

@@ -0,0 +1,3 @@
+import { isBackend } from "./isBackend";
+
+export const isFrontend = !isBackend;