@huggingface/inference - npm Package Compare versions

Comparing version 3.0.0 to 3.0.1

dist/index.js

@@ -52,3 +52,11 @@ var __defProp = Object.defineProperty;
     "black-forest-labs/FLUX.1-schnell": "fal-ai/flux/schnell",
-    "black-forest-labs/FLUX.1-dev": "fal-ai/flux/dev"
+    "black-forest-labs/FLUX.1-dev": "fal-ai/flux/dev",
+    "playgroundai/playground-v2.5-1024px-aesthetic": "fal-ai/playground-v25",
+    "ByteDance/SDXL-Lightning": "fal-ai/lightning-models",
+    "PixArt-alpha/PixArt-Sigma-XL-2-1024-MS": "fal-ai/pixart-sigma",
+    "stabilityai/stable-diffusion-3-medium": "fal-ai/stable-diffusion-v3-medium",
+    "Warlord-K/Sana-1024": "fal-ai/sana",
+    "fal/AuraFlow-v0.2": "fal-ai/aura-flow",
+    "stabilityai/stable-diffusion-3.5-large": "fal-ai/stable-diffusion-v35-large",
+    "Kwai-Kolors/Kolors": "fal-ai/kolors"
   },
@@ -66,6 +74,6 @@ "automatic-speech-recognition": {
     "ByteDance/SDXL-Lightning": "bytedance/sdxl-lightning-4step:5599ed30703defd1d160a25a63321b4dec97101d98b4674bcc56e41f62f35637"
+  },
+  "text-to-speech": {
+    "OuteAI/OuteTTS-0.3-500M": "jbilcke/oute-tts:39a59319327b27327fa3095149c5a746e7f2aee18c75055c3368237a6503cd26"
   }
-  // "text-to-speech": {
-  // 	"SWivid/F5-TTS": "x-lance/f5-tts:87faf6dd7a692dd82043f662e76369cab126a2cf1937e25a9d41e0b834fd230e"
-  // },
 };
@@ -106,2 +114,4 @@
     "databricks/dbrx-instruct": "databricks/dbrx-instruct",
+    "deepseek-ai/DeepSeek-R1": "deepseek-ai/DeepSeek-R1",
+    "deepseek-ai/DeepSeek-V3": "deepseek-ai/DeepSeek-V3",
     "deepseek-ai/deepseek-llm-67b-chat": "deepseek-ai/deepseek-llm-67b-chat",
@@ -152,3 +162,4 @@ "google/gemma-2-9b-it": "google/gemma-2-9b-it",
 async function makeRequestOptions(args, options) {
-  const { accessToken, endpointUrl, provider: maybeProvider, model: maybeModel, ...otherArgs } = args;
+  const { accessToken, endpointUrl, provider: maybeProvider, model: maybeModel, ...remainingArgs } = args;
+  let otherArgs = remainingArgs;
   const provider = maybeProvider ?? "hf-inference";
@@ -212,5 +223,5 @@ const { forceTask, includeCredentials, taskHint, wait_for_model, use_cache, dont_load_model, chatCompletion: chatCompletion2 } = options ?? {};
   }
-  if (provider === "replicate" && model.includes(":")) {
-    const version = model.split(":")[1];
-    otherArgs.version = version;
+  if (provider === "replicate") {
+    const version = model.includes(":") ? model.split(":")[1] : void 0;
+    otherArgs = { input: otherArgs, version };
   }
@@ -588,2 +599,15 @@ const info = {
   });
+  if (res && typeof res === "object") {
+    if ("output" in res) {
+      if (typeof res.output === "string") {
+        const urlResponse = await fetch(res.output);
+        const blob = await urlResponse.blob();
+        return blob;
+      } else if (Array.isArray(res.output)) {
+        const urlResponse = await fetch(res.output[0]);
+        const blob = await urlResponse.blob();
+        return blob;
+      }
+    }
+  }
   const isValidOutput = res && res instanceof Blob;
@@ -670,6 +694,6 @@ if (!isValidOutput) {
     args.prompt = args.inputs;
-    args.inputs = "";
+    delete args.inputs;
     args.response_format = "base64";
   } else if (args.provider === "replicate") {
-    args.input = { prompt: args.inputs };
+    args.prompt = args.inputs;
     delete args.inputs;
@@ -676,0 +700,0 @@ }

package.json

 {
   "name": "@huggingface/inference",
-  "version": "3.0.0",
+  "version": "3.0.1",
   "packageManager": "pnpm@8.10.5",
@@ -42,3 +42,3 @@ "license": "MIT",
   "dependencies": {
-    "@huggingface/tasks": "^0.13.16"
+    "@huggingface/tasks": "^0.13.17"
   },
@@ -45,0 +45,0 @@ "devDependencies": {

README.md

@@ -45,11 +45,11 @@ # 🤗 Hugging Face Inference Endpoints
 
-### Requesting third-party inference providers
+### Third-party inference providers
 
-You can request inference from third-party providers with the inference client.
+You can send inference requests to third-party providers with the inference client.
 
 Currently, we support the following providers: [Fal.ai](https://fal.ai), [Replicate](https://replicate.com), [Together](https://together.xyz) and [Sambanova](https://sambanova.ai).
 
-To make request to a third-party provider, you have to pass the `provider` parameter to the inference function. Make sure your request is authenticated with an access token.
+To send requests to a third-party provider, you have to pass the `provider` parameter to the inference function. Make sure your request is authenticated with an access token.
 ```ts
-const accessToken = "hf_..."; // Either a HF access token, or an API key from the 3rd party provider (Replicate in this example)
+const accessToken = "hf_..."; // Either a HF access token, or an API key from the third-party provider (Replicate in this example)
 
@@ -67,3 +67,3 @@ const client = new HfInference(accessToken);
 
-Only a subset of models are supported when requesting 3rd party providers. You can check the list of supported models per pipeline tasks here:
+Only a subset of models are supported when requesting third-party providers. You can check the list of supported models per pipeline tasks here:
 - [Fal.ai supported models](./src/providers/fal-ai.ts)
@@ -75,4 +75,9 @@ - [Replicate supported models](./src/providers/replicate.ts)
 
-#### Tree-shaking
+❗**Important note:** To be compatible, the third-party API must adhere to the "standard" shape API we expect on HF model pages for each pipeline task type. 
+This is not an issue for LLMs as everyone converged on the OpenAI API anyways, but can be more tricky for other tasks like "text-to-image" or "automatic-speech-recognition" where there exists no standard API. Let us know if any help is needed or if we can make things easier for you!
 
+👋**Want to add another provider?** Get in touch if you'd like to add support for another Inference provider, and/or request it on https://huggingface.co/spaces/huggingface/HuggingDiscussions/discussions/49
+
+### Tree-shaking
+
 You can import the functions you need directly from the module instead of using the `HfInference` class.
@@ -79,0 +84,0 @@

src/lib/makeRequestOptions.ts

@@ -35,3 +35,4 @@ import type { WidgetType } from "@huggingface/tasks";
 ): Promise<{ url: string; info: RequestInit }> {
-	const { accessToken, endpointUrl, provider: maybeProvider, model: maybeModel, ...otherArgs } = args;
+	const { accessToken, endpointUrl, provider: maybeProvider, model: maybeModel, ...remainingArgs } = args;
+	let otherArgs = remainingArgs;
 	const provider = maybeProvider ?? "hf-inference";
@@ -124,8 +125,9 @@
 
-	/*
+	/**
+	 * Replicate models wrap all inputs inside { input: ... }
 	 * Versioned Replicate models in the format `owner/model:version` expect the version in the body
 	 */
-	if (provider === "replicate" && model.includes(":")) {
-		const version = model.split(":")[1];
-		(otherArgs as typeof otherArgs & { version: string }).version = version;
+	if (provider === "replicate") {
+		const version = model.includes(":") ? model.split(":")[1] : undefined;
+		(otherArgs as unknown) = { input: otherArgs, version };
 	}
@@ -132,0 +134,0 @@

src/providers/fal-ai.ts

@@ -11,2 +11,10 @@ import type { ProviderMapping } from "./types";
 		"black-forest-labs/FLUX.1-dev": "fal-ai/flux/dev",
+		"playgroundai/playground-v2.5-1024px-aesthetic": "fal-ai/playground-v25",
+		"ByteDance/SDXL-Lightning": "fal-ai/lightning-models",
+		"PixArt-alpha/PixArt-Sigma-XL-2-1024-MS": "fal-ai/pixart-sigma",
+		"stabilityai/stable-diffusion-3-medium": "fal-ai/stable-diffusion-v3-medium",
+		"Warlord-K/Sana-1024": "fal-ai/sana",
+		"fal/AuraFlow-v0.2": "fal-ai/aura-flow",
+		"stabilityai/stable-diffusion-3.5-large": "fal-ai/stable-diffusion-v35-large",
+		"Kwai-Kolors/Kolors": "fal-ai/kolors",
 	},
@@ -13,0 +21,0 @@ "automatic-speech-recognition": {

src/providers/replicate.ts

@@ -13,5 +13,5 @@ import type { ProviderMapping } from "./types";
 	},
-	// "text-to-speech": {
-	// 	"SWivid/F5-TTS": "x-lance/f5-tts:87faf6dd7a692dd82043f662e76369cab126a2cf1937e25a9d41e0b834fd230e"
-	// },
+	"text-to-speech": {
+		"OuteAI/OuteTTS-0.3-500M": "jbilcke/oute-tts:39a59319327b27327fa3095149c5a746e7f2aee18c75055c3368237a6503cd26",
+	},
 };

src/providers/together.ts

@@ -24,2 +24,4 @@ import type { ProviderMapping } from "./types";
 		"databricks/dbrx-instruct": "databricks/dbrx-instruct",
+		"deepseek-ai/DeepSeek-R1": "deepseek-ai/DeepSeek-R1",
+		"deepseek-ai/DeepSeek-V3": "deepseek-ai/DeepSeek-V3",
 		"deepseek-ai/deepseek-llm-67b-chat": "deepseek-ai/deepseek-llm-67b-chat",
@@ -26,0 +28,0 @@ "google/gemma-2-9b-it": "google/gemma-2-9b-it",

src/tasks/audio/textToSpeech.ts

@@ -13,3 +13,5 @@ import { InferenceOutputError } from "../../lib/InferenceOutputError";
 export type TextToSpeechOutput = Blob;
-
+interface OutputUrlTextToSpeechGeneration {
+	output: string | string[];
+}
 /**
@@ -20,6 +22,19 @@ * This task synthesize an audio of a voice pronouncing a given text.
 export async function textToSpeech(args: TextToSpeechArgs, options?: Options): Promise<TextToSpeechOutput> {
-	const res = await request<TextToSpeechOutput>(args, {
+	const res = await request<TextToSpeechOutput | OutputUrlTextToSpeechGeneration>(args, {
 		...options,
 		taskHint: "text-to-speech",
 	});
+	if (res && typeof res === "object") {
+		if ("output" in res) {
+			if (typeof res.output === "string") {
+				const urlResponse = await fetch(res.output);
+				const blob = await urlResponse.blob();
+				return blob;
+			} else if (Array.isArray(res.output)) {
+				const urlResponse = await fetch(res.output[0]);
+				const blob = await urlResponse.blob();
+				return blob;
+			}
+		}
+	}
 	const isValidOutput = res && res instanceof Blob;
@@ -26,0 +41,0 @@ if (!isValidOutput) {

src/tasks/cv/textToImage.ts

@@ -62,6 +62,6 @@ import { InferenceOutputError } from "../../lib/InferenceOutputError";
 		args.prompt = args.inputs;
-		args.inputs = "";
+		delete (args as unknown as { inputs: unknown }).inputs;
 		args.response_format = "base64";
 	} else if (args.provider === "replicate") {
-		args.input = { prompt: args.inputs };
+		args.prompt = args.inputs;
 		delete (args as unknown as { inputs: unknown }).inputs;
@@ -68,0 +68,0 @@ }

No alert changes

Improved metrics

Total package byte prevSize

284833

increased by1.44%

Lines of code

6236

increased by1.23%

Number of lines in readme file

662

increased by0.76%

Worsened metrics

Number of medium supply chain risk alerts

18

increased by28.57%

Dependency changes

• Updated@huggingface/tasks@^0.13.17