Hugging Face 在 SIGGRAPH 2024 上推出由 NVIDIA NIM 提供支援的推理即服務

開源平台 Hugging Face 正在為開發人員提供由 NVIDIA 的 NIM 提供支援的推理即服務。新服務為 AI 模型提供了 5 倍的令牌效率，並允許立即存取在 NVIDIA DGX Cloud 上運行的 NIM 微服務。

Open Source platform Hugging Face is now offering developers Inference-as-a-Service that will be powered by NVIDIA’s NIM. The new service provides 5x better token efficiency for AI models and allows immediate access to NIM microservices running on NVIDIA DGX Cloud.

開源平台 Hugging Face 現已向開發人員提供由 NVIDIA 的 NIM 提供支援的推理即服務。新服務為 AI 模型提供了 5 倍的令牌效率，並允許立即存取在 NVIDIA DGX Cloud 上運行的 NIM 微服務。

The new inference-as-a-service was announced at the ongoing SIGGRAPH 2024, a premier conference and exhibition on computer graphics and interactive techniques in Denver, Colorado. The new service will let developers deploy powerful LLMs like Llama 2, Mistral AI models and many more with optimisation from NVIDIA NIM microservices. Hugging Face Enterprise Hub users can access serverless inference for increased flexibility and minimal infrastructure overhead with NVIDIA NIM.

新的推理即服務是在科羅拉多州丹佛市正在舉行的 SIGGRAPH 2024 上宣布的，這是一場關於電腦圖形和互動技術的頂級會議和展覽。這項新服務將讓開發人員透過 NVIDIA NIM 微服務的最佳化來部署強大的 LLM，例如 Llama 2、Mistral AI 模型等。 Hugging Face Enterprise Hub 使用者可以使用 NVIDIA NIM 存取無伺服器推理，以提高靈活性並最大限度地減少基礎架構開銷。

When accessed as a NIM, large models like the 70-billion-parameter version of Llama 3 will deliver up to 5x higher throughput when compared with off-the-shelf deployment on NVIDIA H100 Tensor Core GPU-powered systems.

當作為 NIM 存取時，與 NVIDIA H100 Tensor Core GPU 驅動的系統上的現成部署相比，像 Llama 3 的 700 億參數版本這樣的大型模型將提供高達 5 倍的吞吐量。

The new inference service also supports Train on DGX Cloud, an AI training service that is already available on Hugging Face.

新的推理服務還支援 DGX Cloud 上的 Train，這是一項已經在 Hugging Face 上提供的人工智慧訓練服務。

Enter NVIDIA NIM

進入 NVIDIA NIM

NVIDIA NIM is a set of AI microservices, including NVIDIA AI foundation models and open-source community models, that has been optimised for inference with standard APIs. It improves token processing efficiency and enhances the NVIDIA DGX Cloud infrastructure, accelerating AI applications. This setup provides faster, more robust results.

NVIDIA NIM 是一組 AI 微服務，包括 NVIDIA AI 基礎模型和開源社群模型，已針對標準 API 的推理進行了最佳化。它提高了令牌處理效率並增強了 NVIDIA DGX Cloud 基礎設施，從而加速了 AI 應用。此設定可提供更快、更穩健的結果。

The NVIDIA DGX Cloud platform is tailored for generative AI, offering developers reliable, accelerated computing infrastructure for faster production readiness. It supports AI development from prototyping to production without requiring long-term commitments.

NVIDIA DGX 雲端平台專為生成式 AI 量身定制，為開發人員提供可靠、加速的運算基礎設施，以實現更快的生產準備。它支援從原型設計到生產的人工智慧開發，無需長期承諾。

Hugging Face to the Fore

擁抱前面的臉

The new announcement banks on an existing partnership between both tech companies and is only going to foster the developer community further. Interesting recent announcements from Hugging Face include its profitability with a 220-member team and the release of SmolLM, a series of small language models.

新的公告是基於兩家科技公司之間現有的合作夥伴關係，只會進一步培養開發者社群。 Hugging Face 最近發布的有趣消息包括其擁有 220 名成員的團隊實現盈利，以及發布 SmolLM（一系列小語言模型）。