File Manager - BorneoXploit

Editing: app_llm2.html

<!DOCTYPE html>
<html class="writer-html5" lang="en" data-content_root="../">
<head>
  <meta charset="utf-8" />
  <meta name="readthedocs-addons-api-version" content="1"><meta name="viewport" content="width=device-width, initial-scale=1" />

<a href="../contents.html">
            
              <img src="../_static/logo-white.png" class="logo" alt="Logo"/>
          </a>
              <div class="switch-menus">
                <div class="version-switch"></div>
                <div class="language-switch"></div>
              </div>
<div role="search">
  <form id="rtd-search-form" class="wy-form" action="../search.html" method="get">
    <input type="text" name="q" placeholder="Search docs" aria-label="Search docs" />
    <input type="hidden" name="check_keywords" value="yes" />
    <input type="hidden" name="area" value="default" />
  </form>
</div>
        </div><div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="Navigation menu">
              <p class="caption" role="heading"><span class="caption-text">Getting Started</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../index.html">Introduction</a></li>
<li class="toctree-l1"><a class="reference internal" href="../release_schedule.html">Maintenance and release schedule</a></li>
<li class="toctree-l1"><a class="reference internal" href="../gdpr/index.html">GDPR-compliance</a></li>
<li class="toctree-l1"><a class="reference internal" href="../declarations/index.html">Declarations</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Release notes</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../release_notes/index.html">Critical changes</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Installation</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../installation/index.html">Installation and server configuration</a></li>
<li class="toctree-l1"><a class="reference internal" href="../configuration_database/index.html">Database configuration</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Configuration</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../configuration_server/index.html">Nextcloud configuration</a></li>
<li class="toctree-l1"><a class="reference internal" href="../occ_command.html">Using the occ command</a></li>
<li class="toctree-l1"><a class="reference internal" href="../reference/index.html">Reference management</a></li>
<li class="toctree-l1"><a class="reference internal" href="../webhook_listeners/index.html">Webhook Listeners</a></li>
<li class="toctree-l1"><a class="reference internal" href="../windmill_workflows/index.html">Windmill Workflows</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Files</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../configuration_files/index.html">File sharing and management</a></li>
<li class="toctree-l1"><a class="reference internal" href="../file_workflows/index.html">Flow</a></li>
<li class="toctree-l1"><a class="reference internal" href="../configuration_mimetypes/index.html">Mimetypes management</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Apps</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../apps_management.html">Apps management</a></li>
<li class="toctree-l1"><a class="reference internal" href="../exapps_management/index.html">ExApps management</a></li>
<li class="toctree-l1 current"><a class="reference internal" href="index.html">Artificial Intelligence</a><ul class="current">
<li class="toctree-l2"><a class="reference internal" href="overview.html">Overview</a></li>
<li class="toctree-l2"><a class="reference internal" href="app_assistant.html">Nextcloud Assistant</a></li>
<li class="toctree-l2"><a class="reference internal" href="app_translate2.html">App: Local Machine translation 2 (translate2)</a></li>
<li class="toctree-l2 current"><a class="current reference internal" href="#">App: Local large language model (llm2)</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#multilinguality">Multilinguality</a></li>
<li class="toctree-l3"><a class="reference internal" href="#requirements">Requirements</a></li>
<li class="toctree-l3"><a class="reference internal" href="#installation">Installation</a><ul>
<li class="toctree-l4"><a class="reference internal" href="#supplying-alternate-models">Supplying alternate models</a></li>
<li class="toctree-l4"><a class="reference internal" href="#configuring-alternate-models">Configuring alternate models</a></li>
</ul>
</li>
<li class="toctree-l3"><a class="reference internal" href="#scaling">Scaling</a></li>
<li class="toctree-l3"><a class="reference internal" href="#app-store">App store</a></li>
<li class="toctree-l3"><a class="reference internal" href="#repository">Repository</a></li>
<li class="toctree-l3"><a class="reference internal" href="#known-limitations">Known Limitations</a></li>
<li class="toctree-l3"><a class="reference internal" href="#addendum-running-with-a-fully-open-model">Addendum: Running with a fully open model</a><ul>
<li class="toctree-l4"><a class="reference internal" href="#what-makes-olmo-a-fully-open-model">What makes OLMo a fully open model?</a></li>
<li class="toctree-l4"><a class="reference internal" href="#limitations">Limitations</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="app_stt_whisper2.html">App: Local Whisper Speech-To-Text (stt_whisper2)</a></li>
<li class="toctree-l2"><a class="reference internal" href="app_text2image_stablediffusion2.html">App: Local Image Generation (text2image_stablediffusion2)</a></li>
<li class="toctree-l2"><a class="reference internal" href="app_recognize.html">App: Recognize</a></li>
<li class="toctree-l2"><a class="reference internal" href="app_context_chat.html">App: Context Chat</a></li>
<li class="toctree-l2"><a class="reference internal" href="app_context_agent.html">App: Context Agent (context_agent)</a></li>
<li class="toctree-l2"><a class="reference internal" href="app_summary_bot.html">App: Summary Bot (Talk chat summarize bot)</a></li>
<li class="toctree-l2"><a class="reference internal" href="app_text2speech_kokoro.html">App: Local Text-To-Speech (text2speech_kokoro)</a></li>
<li class="toctree-l2"><a class="reference internal" href="app_live_transcription.html">App: Live Transcription and Translation in Nextcloud Talk (live_transcription)</a></li>
<li class="toctree-l2"><a class="reference internal" href="ai_as_a_service.html">AI as a Service</a></li>
<li class="toctree-l2"><a class="reference internal" href="insight_and_debugging.html">Insight and debugging</a></li>
</ul>
</li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Users</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../configuration_user/index.html">User management</a></li>
<li class="toctree-l1"><a class="reference internal" href="../desktop/index.html">Desktop Clients</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Groupware</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../groupware/index.html">Groupware</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Office</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../office/index.html">Office</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Maintenance</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../configuration_monitoring/index.html">Monitoring</a></li>
<li class="toctree-l1"><a class="reference internal" href="../maintenance/index.html">Maintenance</a></li>
<li class="toctree-l1"><a class="reference internal" href="../issues/index.html">Issues and troubleshooting</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Links</span></p>
<ul>
<li class="toctree-l1"><a class="reference external" href="https://help.nextcloud.com/">Community Help Forum</a></li>
<li class="toctree-l1"><a class="reference external" href="https://docs.nextcloud.com/">User Manuals</a></li>
<li class="toctree-l1"><a class="reference external" href="https://docs.nextcloud.com/">Developer Manuals</a></li>
<li class="toctree-l1"><a class="reference external" href="https://nextcloud.com/install/">Download</a></li>
<li class="toctree-l1"><a class="reference external" href="https://apps.nextcloud.com">App Store</a></li>
<li class="toctree-l1"><a class="reference external" href="https://help.nextcloud.com/t/translation-knowledge-valid-for-the-entire-nextcloud-project-wiki/51550">Translations</a></li>
<li class="toctree-l1"><a class="reference external" href="https://github.com/nextcloud/">GitHub</a></li>
<li class="toctree-l1"><a class="reference external" href="https://github.com/nextcloud/server/wiki/Maintenance-and-Release-Schedule">Supported versions</a></li>
<li class="toctree-l1"><a class="reference external" href="https://nextcloud.com/enterprise/">Nextcloud Enterprise</a></li>
<li class="toctree-l1"><a class="reference external" href="https://github.com/nextcloud/server/blob/master/COPYING-README">License</a></li>
<li class="toctree-l1"><a class="reference external" href="https://nextcloud.com/">Nextcloud GmbH</a></li>
</ul>

</div>
      </div>
    </nav>

<section data-toggle="wy-nav-shift" class="wy-nav-content-wrap"><nav class="wy-nav-top" aria-label="Mobile navigation menu" >
          <i data-toggle="wy-nav-top" class="fa fa-bars"></i>
          <a href="../contents.html">Nextcloud latest Administration Manual</a>
      </nav>

<div class="wy-nav-content">
        <div class="rst-content style-external-links">
          <div role="navigation" aria-label="Page navigation">
  <ul class="wy-breadcrumbs">
      <li><a href="../contents.html" class="icon icon-home" aria-label="Home"></a></li>
          <li class="breadcrumb-item"><a href="index.html">Artificial Intelligence</a></li>
      <li class="breadcrumb-item active">App: Local large language model (llm2)</li>
      <li class="wy-breadcrumbs-aside">
              <a href="https://github.com/nextcloud/documentation/edit/master/admin_manual/ai/app_llm2.rst" class="fa fa-github"> Edit on GitHub</a>
      </li>
  </ul>
  <hr/>
</div>
          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
           <div itemprop="articleBody">
             
  <section id="app-local-large-language-model-llm2">
<h1>App: Local large language model (llm2)<a class="headerlink" href="#app-local-large-language-model-llm2" title="Link to this heading"></a></h1>
<p id="ai-app-llm2">The <em>llm2</em> app is one of the apps that provide text processing functionality using Large language models in Nextcloud and act as a text processing backend for the <a class="reference internal" href="app_assistant.html#ai-app-assistant"><span class="std std-ref">Nextcloud Assistant app</span></a>, the <em>mail</em> app and <a class="reference internal" href="overview.html#tp-consumer-apps"><span class="std std-ref">other apps making use of the core Text Processing API</span></a>. The <em>llm2</em> app specifically runs only open source models and does so entirely on-premises. Nextcloud can provide customer support upon request, please talk to your account manager for the possibilities.</p>
<p>This app uses <a class="reference external" href="https://github.com/abetlen/llama-cpp-python">llama.cpp</a> under the hood and is thus compatible with any model in <em>gguf</em> format.</p>
<p>However, we only test with Llama 3.1. Output quality will differ depending on which model you use and downstream tasks like summarization or Context Chat may not work on other models.
We thus recommend the following models:</p>
<ul class="simple">
<li><p><a class="reference external" href="https://huggingface.co/QuantFactory/Meta-Llama-3.1-8B-Instruct-GGUF">Llama3.1 8b Instruct</a> (reasonable quality; fast; good acclaim; comes shipped with the app)</p></li>
<li><p><a class="reference external" href="https://huggingface.co/bartowski/Meta-Llama-3.1-70B-Instruct-GGUF">Llama3.1 70B Instruct</a> (good quality; good acclaim)</p></li>
</ul>
<section id="multilinguality">
<h2>Multilinguality<a class="headerlink" href="#multilinguality" title="Link to this heading"></a></h2>
<p>This app supports input and output in languages other than English if the underlying model supports the language.</p>
<p>Llama 3.1 <a class="reference external" href="https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct#multilingual-benchmarks">supports the following languages:</a></p>
<ul class="simple">
<li><p>English</p></li>
<li><p>Portuguese</p></li>
<li><p>Spanish</p></li>
<li><p>Italian</p></li>
<li><p>German</p></li>
<li><p>French</p></li>
<li><p>Hindi</p></li>
<li><p>Thai</p></li>
</ul>
<p>Note, that other languages may work as well, but only the above languages are guaranteed to work.</p>
</section>
<section id="requirements">
<h2>Requirements<a class="headerlink" href="#requirements" title="Link to this heading"></a></h2>
<ul>
<li><p>This app is built as an External App and thus depends on AppAPI v3.1.0 or higher</p></li>
<li><p>Nextcloud AIO is supported</p></li>
<li><p>We currently support NVIDIA GPUs and x86_64 CPUs</p></li>
<li><p>CPU that supports AVX and AVX2 instruction</p></li>
<li><p>CUDA &gt;= v12.4 on your host system</p></li>
<li><p>GPU Sizing</p>
<blockquote>
<div><ul class="simple">
<li><p>A NVIDIA GPU with at least 8GB VRAM</p></li>
<li><p>At least 12GB of system RAM</p></li>
</ul>
</div></blockquote>
</li>
<li><p>CPU Sizing</p>
<blockquote>
<div><ul class="simple">
<li><p>At least 12GB of system RAM</p></li>
<li><p>The more cores you have and the more powerful the CPU the better, we recommend 10-20 cores</p></li>
<li><p>The app will hog all cores by default, so it is usually better to run it on a separate machine</p></li>
</ul>
</div></blockquote>
</li>
</ul>
</section>
<section id="installation">
<h2>Installation<a class="headerlink" href="#installation" title="Link to this heading"></a></h2>
<ol class="arabic simple" start="0">
<li><p>Make sure the <a class="reference internal" href="app_assistant.html#ai-app-assistant"><span class="std std-ref">Nextcloud Assistant app</span></a> is installed</p></li>
<li><p><a class="reference internal" href="../exapps_management/AppAPIAndExternalApps.html#ai-app-api"><span class="std std-ref">Install AppAPI and setup a Deploy Demon</span></a></p></li>
<li><p>Install the “Local large language model” ExApp via the “Apps” page in the Nextcloud web admin user interface</p></li>
</ol>
<section id="supplying-alternate-models">
<h3>Supplying alternate models<a class="headerlink" href="#supplying-alternate-models" title="Link to this heading"></a></h3>
<p>This app allows supplying alternate LLM models as <em>gguf</em> files in the <code class="docutils literal notranslate"><span class="pre">/nc_app_llm2_data</span></code> directory of the docker container.</p>
<ol class="arabic simple">
<li><p>Download a <strong>gguf</strong> model e.g. from huggingface</p></li>
<li><p>Copy the <strong>gguf</strong> file to <code class="docutils literal notranslate"><span class="pre">/nc_app_llm2_data</span></code> inside the docker container</p></li>
<li><p>Restart the llm2 ExApp</p></li>
<li><p>Select the new model in the Nextcloud AI admin settings</p></li>
</ol>
</section>
<section id="configuring-alternate-models">
<h3>Configuring alternate models<a class="headerlink" href="#configuring-alternate-models" title="Link to this heading"></a></h3>
<p>Since every model requires slightly different inference parameters, you can pass along a configuration file for the alternate model files you supply.</p>
<p>The configuration file for a model file must have the same name as the model file but must end in <code class="docutils literal notranslate"><span class="pre">.json</span></code> instead of <code class="docutils literal notranslate"><span class="pre">.gguf</span></code>.</p>
<p>The strings <code class="docutils literal notranslate"><span class="pre">{system_prompt}</span></code> and <code class="docutils literal notranslate"><span class="pre">{user_prompt}</span></code> are variables that will be filled in by the app, so they must be part of your prompt template.</p>
<p>Here is an example config file for Llama 2:</p>
<div class="highlight-json notranslate"><div class="highlight"><pre><span></span><span class="p">{</span>
<span class="w">  </span><span class="nt">&quot;prompt&quot;</span><span class="p">:</span><span class="w"> </span><span class="s2">&quot;&lt;|im_start|&gt; system\n{system_prompt}\n&lt;|im_end|&gt;\n&lt;|im_start|&gt; user\n{user_prompt}\n&lt;|im_end|&gt;\n&lt;|im_start|&gt; assistant\n&quot;</span><span class="p">,</span>
<span class="w">  </span><span class="nt">&quot;loader_config&quot;</span><span class="p">:</span><span class="w"> </span><span class="p">{</span>
<span class="w">     </span><span class="nt">&quot;n_ctx&quot;</span><span class="p">:</span><span class="w"> </span><span class="mi">4096</span><span class="p">,</span>
<span class="w">     </span><span class="nt">&quot;max_tokens&quot;</span><span class="p">:</span><span class="w"> </span><span class="mi">2048</span><span class="p">,</span>
<span class="w">     </span><span class="nt">&quot;stop&quot;</span><span class="p">:</span><span class="w"> </span><span class="p">[</span><span class="s2">&quot;&lt;|im_end|&gt;&quot;</span><span class="p">]</span>
<span class="w">  </span><span class="p">}</span>
<span class="p">}</span>
</pre></div>
</div>
<p>Here is an example configuration for Llama 3:</p>
<div class="highlight-json notranslate"><div class="highlight"><pre><span></span><span class="p">{</span>
<span class="w">  </span><span class="nt">&quot;prompt&quot;</span><span class="p">:</span><span class="w"> </span><span class="s2">&quot;&lt;|begin_of_text|&gt;&lt;|start_header_id|&gt;system&lt;|end_header_id|&gt;\n{system_prompt}&lt;|eot_id|&gt;&lt;|start_header_id|&gt;user&lt;|end_header_id|&gt;\n{user_prompt}&lt;|eot_id|&gt;\n&lt;|start_header_id|&gt;assistant&lt;|end_header_id|&gt;\n&quot;</span><span class="p">,</span>
<span class="w">  </span><span class="nt">&quot;loader_config&quot;</span><span class="p">:</span><span class="w"> </span><span class="p">{</span>
<span class="w">      </span><span class="nt">&quot;n_ctx&quot;</span><span class="p">:</span><span class="w"> </span><span class="mi">8000</span><span class="p">,</span>
<span class="w">      </span><span class="nt">&quot;max_tokens&quot;</span><span class="p">:</span><span class="w"> </span><span class="mi">4000</span><span class="p">,</span>
<span class="w">      </span><span class="nt">&quot;stop&quot;</span><span class="p">:</span><span class="w"> </span><span class="p">[</span><span class="s2">&quot;&lt;|eot_id|&gt;&quot;</span><span class="p">],</span>
<span class="w">      </span><span class="nt">&quot;temperature&quot;</span><span class="p">:</span><span class="w"> </span><span class="mf">0.3</span>
<span class="w">  </span><span class="p">}</span>
<span class="p">}</span>
</pre></div>
</div>
</section>
</section>
<section id="scaling">
<h2>Scaling<a class="headerlink" href="#scaling" title="Link to this heading"></a></h2>
<p>It is currently not possible to scale this app, we are working on this. Based on our calculations an instance has a rough capacity of 1000 user requests per hour. However, this number is based on theory and we do appreciate real-world feedback on this.
If you would like to scale up your language model usage, we recommend using an <a class="reference internal" href="ai_as_a_service.html#ai-ai-as-a-service"><span class="std std-ref">AI as a Service provider</span></a> or hosting a service compatible with the OpenAI API yourself that can be scaled up and connecting nextcloud to it via the <a class="reference external" href="https://apps.nextcloud.com/apps/integration_openai">integration_openai app</a>.</p>
</section>
<section id="app-store">
<h2>App store<a class="headerlink" href="#app-store" title="Link to this heading"></a></h2>
<p>You can also find the app in our app store, where you can write a review: <a class="reference external" href="https://apps.nextcloud.com/apps/llm2">https://apps.nextcloud.com/apps/llm2</a></p>
</section>
<section id="repository">
<h2>Repository<a class="headerlink" href="#repository" title="Link to this heading"></a></h2>
<p>You can find the app’s code repository on GitHub where you can report bugs and contribute fixes and features: <a class="reference external" href="https://github.com/nextcloud/llm2">https://github.com/nextcloud/llm2</a></p>
<p>Nextcloud customers should file bugs directly with our Support system.</p>
</section>
<section id="known-limitations">
<h2>Known Limitations<a class="headerlink" href="#known-limitations" title="Link to this heading"></a></h2>
<ul class="simple">
<li><p>We currently only support languages that the underlying model supports; correctness of language use in languages other than English may be poor depending on the language’s coverage in the model’s training data (We recommended model Llama 3 or other models explicitly trained on multiple languages)</p></li>
<li><p>Language models can be bad at reasoning tasks</p></li>
<li><p>Language models can be bad at math</p></li>
<li><p>Language models are likely to generate false information and should thus only be used in situations that are not critical. It’s recommended to only use AI at the beginning of a creation process and not at the end, so that outputs of AI serve as a draft for example and not as final product. Always check the output of language models before using it.</p></li>
<li><p>Make sure to test the language model you are using it for whether it meets the use-case’s quality requirements</p></li>
<li><p>Language models notoriously have a high energy consumption, if you want to reduce load on your server you can choose smaller models or quantized models in exchange for lower accuracy</p></li>
<li><p>Customer support is available upon request, however we can’t solve false or problematic output, most performance issues, or other problems caused by the underlying model. Support is thus limited only to bugs directly caused by the implementation of the app (connectors, API, front-end, AppAPI)</p></li>
</ul>
</section>
<section id="addendum-running-with-a-fully-open-model">
<h2>Addendum: Running with a fully open model<a class="headerlink" href="#addendum-running-with-a-fully-open-model" title="Link to this heading"></a></h2>
<p>If you would like to use a fully open model that scores a green score on our Ethical AI rating, we recommend the following model:</p>
<ul class="simple">
<li><p>Olmo 3 (either in 7B or 32B): <a class="reference external" href="https://huggingface.co/allenai/Olmo-3-7B-Instruct">https://huggingface.co/allenai/Olmo-3-7B-Instruct</a></p></li>
</ul>
<section id="what-makes-olmo-a-fully-open-model">
<h3>What makes OLMo a fully open model?<a class="headerlink" href="#what-makes-olmo-a-fully-open-model" title="Link to this heading"></a></h3>
<ul class="simple">
<li><p>The code for training, fine-tuning and inference of the model is publicly available and fully open source</p></li>
<li><p>The training data with which the model is pretrained is publicly available</p></li>
<li><p>The model itself is publicly available and fully open source</p></li>
<li><p>The instruction tuning data is publicly available</p></li>
<li><p>The Reinforcement learning model is publicly available and fully open source</p></li>
</ul>
</section>
<section id="limitations">
<h3>Limitations<a class="headerlink" href="#limitations" title="Link to this heading"></a></h3>
<ul class="simple">
<li><p>OLMo currently only works well with English language input</p></li>
<li><p>In our tests it sometimes produced hallucinated or garbled output; make sure to thoroughly test the model for your use case</p></li>
<li><p>It cannot use tools, so cannot be used in conjunction with <a class="reference internal" href="app_context_agent.html#ai-app-context-agent"><span class="std std-ref">Context Agent</span></a></p></li>
</ul>
</section>
</section>
</section>

</div>
          </div>
          <footer><div class="rst-footer-buttons" role="navigation" aria-label="Footer">
        <a href="app_translate2.html" class="btn btn-neutral float-left" title="App: Local Machine translation 2 (translate2)" accesskey="p" rel="prev"><span class="fa fa-arrow-circle-left" aria-hidden="true"></span> Previous</a>
        <a href="app_stt_whisper2.html" class="btn btn-neutral float-right" title="App: Local Whisper Speech-To-Text (stt_whisper2)" accesskey="n" rel="next">Next <span class="fa fa-arrow-circle-right" aria-hidden="true"></span></a>
    </div>

<hr/>

<div role="contentinfo">
    <p>&#169; Copyright 2016-2026 Nextcloud GmbH and Nextcloud contributors.</p>
  </div>

</footer>
        </div>
      </div>
    </section>
  </div>

<div class="rst-versions" data-toggle="rst-versions" role="note" aria-label="versions">
    <span class="rst-current-version" data-toggle="rst-current-version">
      ☁️ latest
      <span class="fa fa-caret-down"></span>
    </span>
    <div class="rst-other-versions">
      <dl>
        <dt>☁️ Versions</dt>
        
        <dd style="width: 32%">
          <a href="https://docs.nextcloud.com/server/latest/admin_manual"
             
             style="color: var(--dark-link-color);"
             
          >
            latest
          </a>
        </dd>
        
        <dd style="width: 32%">
          <a href="https://docs.nextcloud.com/server/stable/admin_manual"
             
          >
            stable
          </a>
        </dd>
        
        <dd style="width: 32%">
          <a href="https://docs.nextcloud.com/server/32/admin_manual"
             
          >
            32
          </a>
        </dd>
        
        <dd style="width: 32%">
          <a href="https://docs.nextcloud.com/server/31/admin_manual"
             
          >
            31
          </a>
        </dd>
        
      </dl>
    </div>
  </div>
<script>
      jQuery(function () {
          SphinxRtdTheme.Navigation.enable(true);
      });
  </script>

</body>
</html>

Cancel