Meta 推出多標記預測技術，可能徹底改變大型語言模型開發

Meta 在更有效率的人工智慧競賽中發起了挑戰。這家科技巨頭週三發布了預訓練模型，該模型利用了一種新穎的多令牌預測方法，可能會改變大型語言模型（LLM）的開發和部署方式。

Meta unveiled pre-trained models on Wednesday that leverage a novel multi-token prediction approach, potentially changing how large language models (LLMs) are developed and deployed.

Meta 於週三推出了預訓練模型，該模型利用了一種新穎的多標記預測方法，可能會改變大型語言模型 (LLM) 的開發和部署方式。

The tech giant’s latest offering comes in the wake of a recent paper published by Meta researchers, which outlines a new training method for LLMs that leverages multi-token prediction. In a bid to further propel research in this domain, Meta has now released pre-trained models for code completion, leveraging this approach on Hugging Face.

這家科技巨頭的最新產品是在 Meta 研究人員最近發表的一篇論文之後推出的，該論文概述了一種利用多令牌預測的 LLM 的新訓練方法。為了進一步推動這一領域的研究，Meta 現在發布了用於程式碼完成的預訓練模型，在 Hugging Face 上利用了這種方法。

This technique marks a departure from the traditional approach of training LLMs to predict only the next word in a sequence. Instead, Meta’s method tasks models with forecasting multiple future words simultaneously, promising both enhanced performance and drastically reduced training times.

這項技術標誌著與訓練法學碩士僅預測序列中下一個單字的傳統方法的背離。相反，Meta 的方法任務模型同時預測多個未來單詞，有望提高表現並大幅減少訓練時間。

The implications of this breakthrough could be far-reaching. As AI models continue to grow in size and complexity, their voracious appetite for computational power has raised concerns about cost and environmental impact. Meta’s multi-token prediction method might offer a way to curb this trend, making advanced AI more accessible and sustainable.

這一突破的影響可能是深遠的。隨著人工智慧模型的規模和複雜性不斷增長，它們對運算能力的貪婪需求引起了人們對成本和環境影響的擔憂。 Meta 的多代幣預測方法可能提供一種遏制這種趨勢的方法，使先進的人工智慧更容易獲得和可持續。

Democratizing AI: The promise and perils of efficient language models

人工智慧民主化：高效語言模型的前景與危險

The potential of this new approach extends beyond mere efficiency gains. By predicting multiple tokens at once, these models may develop a more nuanced understanding of language structure and context. This could lead to improvements in tasks ranging from code generation to creative writing, potentially bridging the gap between AI and human-level language understanding.

這種新方法的潛力不僅僅是提高效率。透過同時預測多個標記，這些模型可以對語言結構和上下文產生更細緻的理解。這可能會導致從程式碼生成到創意寫作等任務的改進，有可能縮小人工智慧和人類語言理解之間的差距。

Countdown to VB Transform 2024

VB 轉型 2024 倒數計時

Join enterprise leaders in San Francisco from July 9 to 11 for our flagship AI event. Connect with peers, explore the opportunities and challenges of Generative AI, and learn how to integrate AI applications into your industry. Register Now

7 月 9 日至 11 日，與舊金山的企業領袖一起參加我們的旗艦人工智慧活動。與同行交流，探索生成式人工智慧的機會和挑戰，並了解如何將人工智慧應用程式整合到您的行業中。現在註冊