---
title: "#directpreferenceoptimization articles | Hashnode"
description: "Explore 2 articles tagged #directpreferenceoptimization on Hashnode. Tutorials, guides, and how-tos from the community."
url: "https://hashnode.com/tag/directpreferenceoptimization"
---

Tag

# #directpreferenceoptimization

2 posts · 0 followers

[Jonas Kim](https://hashnode.com/@bits-bytes-nn) · [Jan 7, 2024](https://hashnode.com/posts/mixtral-of-experts/67f3c2aba3ed3a26867b04f1)

### [Mixtral of Experts 논문 리뷰](https://bits-bytes-nn.hashnode.dev/mixtral-of-experts?utm_source=hashnode&utm_medium=feed)

Mixtral은 오픈소스 LLM에 전문가 혼합(MoE) 구조를 도입하여, 추론 속도와 처리량을 크게 향상시켰습니다.

24 min read · [bits-bytes-nn.hashnode.dev](https://bits-bytes-nn.hashnode.dev/mixtral-of-experts?utm_source=hashnode&utm_medium=feed)

[Jonas Kim](https://hashnode.com/@bits-bytes-nn) · [Jan 4, 2024](https://hashnode.com/posts/deepseek-llm-scaling-open-source-language-models-with-longtermism/67e2c222dc0e88ba125abf9f)

### [DeepSeek LLM: Scaling Open-Source Language Models with Longtermism 논문 리뷰](https://bits-bytes-nn.hashnode.dev/deepseek-llm-scaling-open-source-language-models-with-longtermism?utm_source=hashnode&utm_medium=feed)

DeepSeek AI는 영어와 중국어 분야에서 우수한 성능을 보이는 모델을 선보이는 동시에, 하이퍼파라미터와 모델·데이터 간 연산량 할당에 관한 정교한 스케일링 법칙도 제시했습니다.

21 min read · [bits-bytes-nn.hashnode.dev](https://bits-bytes-nn.hashnode.dev/deepseek-llm-scaling-open-source-language-models-with-longtermism?utm_source=hashnode&utm_medium=feed)
