<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>NLP on Coriva</title>
    <link>https://coriva.eu.org/categories/nlp/</link>
    <description>Recent content in NLP on Coriva</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Fri, 03 Jan 2025 11:34:14 +0800</lastBuildDate>
    <atom:link href="https://coriva.eu.org/categories/nlp/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>MEDUSA: Detailed Explanation of the Mechanism</title>
      <link>https://coriva.eu.org/medusa-detailed-explanation-of-the-mechanism/</link>
      <pubDate>Fri, 03 Jan 2025 11:34:14 +0800</pubDate>
      <guid>https://coriva.eu.org/medusa-detailed-explanation-of-the-mechanism/</guid>
      <description>A deep dive into MEDUSA, an LLM inference acceleration framework using multiple decoding heads for faster text generation.</description>
    </item>
    <item>
      <title>大模型创业前必须想清楚的问题</title>
      <link>https://coriva.eu.org/llm-based-startup-issues/</link>
      <pubDate>Thu, 21 Dec 2023 15:43:31 +0800</pubDate>
      <guid>https://coriva.eu.org/llm-based-startup-issues/</guid>
      <description>大模型创业前需要想清楚的关键问题：技术壁垒、商业模式与竞争格局。</description>
    </item>
    <item>
      <title>Phi-2: 小型语言模型的惊人威力</title>
      <link>https://coriva.eu.org/phi-2-summary/</link>
      <pubDate>Thu, 14 Dec 2023 17:34:37 +0800</pubDate>
      <guid>https://coriva.eu.org/phi-2-summary/</guid>
      <description>微软Phi-2小型语言模型的技术解析：27亿参数如何媲美更大模型的性能。</description>
    </item>
    <item>
      <title>Textbooks Are All You Need 精要</title>
      <link>https://coriva.eu.org/textbooks-are-all-you-need-summary/</link>
      <pubDate>Wed, 13 Dec 2023 00:33:39 +0800</pubDate>
      <guid>https://coriva.eu.org/textbooks-are-all-you-need-summary/</guid>
      <description>微软Phi-1论文精要：高质量教科书数据如何让小模型实现大模型的能力。</description>
    </item>
    <item>
      <title>检索增强生成（RAG）简介</title>
      <link>https://coriva.eu.org/intro-to-rag/</link>
      <pubDate>Wed, 06 Dec 2023 23:50:54 +0800</pubDate>
      <guid>https://coriva.eu.org/intro-to-rag/</guid>
      <description>检索增强生成（RAG）技术的原理介绍，结合检索与大语言模型提升回答质量。</description>
    </item>
  </channel>
</rss>
