<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Nlp on Le Site de François</title><link>https://lesitedefrancois.be/en/tags/nlp/</link><description>Recent content in Nlp on Le Site de François</description><generator>Hugo -- gohugo.io</generator><language>en</language><copyright>© 2026 François</copyright><atom:link href="https://lesitedefrancois.be/en/tags/nlp/index.xml" rel="self" type="application/rss+xml"/><item><title>LLM (Large Language Model)</title><link>https://lesitedefrancois.be/en/ai/llm/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://lesitedefrancois.be/en/ai/llm/</guid><description>&lt;p&gt;An &lt;strong&gt;LLM (large language model)&lt;/strong&gt; is a deep neural network—almost always a &lt;strong&gt;Transformer&lt;/strong&gt;—trained on large amounts of text (and sometimes multimodal data) to model the probability of the next token given prior context. Its objective at &lt;strong&gt;training&lt;/strong&gt; time is to minimize prediction error over billions of tokens, producing weights that encode grammar, facts (with limitations), reasoning patterns, and task-following behavior after alignment or instruction tuning. At &lt;strong&gt;inference&lt;/strong&gt; time the same model &lt;strong&gt;generates&lt;/strong&gt; completions, answers questions, summarizes documents, or drives agents; production systems expose it through APIs (often OpenAI-compatible) backed by engines such as &lt;strong&gt;vLLM&lt;/strong&gt; or &lt;strong&gt;NIM&lt;/strong&gt;. LLMs power chatbots, code assistants, RAG pipelines, and enterprise copilots.&lt;/p&gt;</description></item></channel></rss>