GitHub/ReAgent
1
0
Fork 0
mirror of https://github.com/facebookresearch/ReAgent.git synced 2026-06-16 12:44:41 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
gh-pages
ReAgent/api/reagent.data.html
T
Zhengxing Chen f2d2a25aff Update docs 11/20/2021
2021-11-20 20:47:01 -08:00

1099 lines
148 KiB
HTML
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
<!DOCTYPE html>
<html class="writer-html5" lang="en" >
<head>
<meta charset="utf-8" /><meta name="generator" content="Docutils 0.17.1: http://docutils.sourceforge.net/" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>reagent.data package &mdash; ReAgent 1.0 documentation</title>
<link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
<link rel="stylesheet" href="../_static/css/theme.css" type="text/css" />
<!--[if lt IE 9]>
<script src="../_static/js/html5shiv.min.js"></script>
<![endif]-->
<script data-url_root="../" id="documentation_options" src="../_static/documentation_options.js"></script>
<script src="../_static/jquery.js"></script>
<script src="../_static/underscore.js"></script>
<script src="../_static/doctools.js"></script>
<script src="../_static/js/theme.js"></script>
<link rel="index" title="Index" href="../genindex.html" />
<link rel="search" title="Search" href="../search.html" />
<link rel="next" title="reagent.gym package" href="reagent.gym.html" />
<link rel="prev" title="reagent.core package" href="reagent.core.html" />
</head>
<body class="wy-body-for-nav">
<div class="wy-grid-for-nav">
<nav data-toggle="wy-nav-shift" class="wy-nav-side">
<div class="wy-side-scroll">
<div class="wy-side-nav-search" >
<a href="../index.html" class="icon icon-home"> ReAgent
</a>
<div role="search">
<form id="rtd-search-form" class="wy-form" action="../search.html" method="get">
<input type="text" name="q" placeholder="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="../installation.html">Installation</a></li>
<li class="toctree-l1"><a class="reference internal" href="../usage.html">Usage</a></li>
<li class="toctree-l1"><a class="reference internal" href="../rasp_tutorial.html">RASP (Not Actively Maintained)</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Advanced Topics</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../continuous_integration.html">Continuous Integration</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Package Reference</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="reagent.core.html">Core</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">Data</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#submodules">Submodules</a></li>
<li class="toctree-l2"><a class="reference internal" href="#module-reagent.data.data_fetcher">reagent.data.data_fetcher module</a></li>
<li class="toctree-l2"><a class="reference internal" href="#module-reagent.data.manual_data_module">reagent.data.manual_data_module module</a></li>
<li class="toctree-l2"><a class="reference internal" href="#module-reagent.data.oss_data_fetcher">reagent.data.oss_data_fetcher module</a></li>
<li class="toctree-l2"><a class="reference internal" href="#module-reagent.data.reagent_data_module">reagent.data.reagent_data_module module</a></li>
<li class="toctree-l2"><a class="reference internal" href="#module-reagent.data.spark_utils">reagent.data.spark_utils module</a></li>
<li class="toctree-l2"><a class="reference internal" href="#module-reagent.data">Module contents</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="reagent.gym.html">Gym</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.evaluation.html">Evaluation</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.lite.html">Lite</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.mab.html">MAB</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.model_managers.html">Model Managers</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.model_utils.html">Model Utils</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.net_builder.html">Net Builders</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.optimizer.html">Optimizers</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.models.html">Models</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.prediction.html">Prediction</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.preprocessing.html">Preprocessing</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.training.html">Training</a></li>
<li class="toctree-l1"><a class="reference internal" href="reagent.workflow.html">Workflow</a></li>
<li class="toctree-l1"><a class="reference internal" href="modules.html">All Modules</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Others</span></p>
<ul>
<li class="toctree-l1"><a class="reference external" href="https://github.com/facebookresearch/ReAgent">Github</a></li>
<li class="toctree-l1"><a class="reference internal" href="../license.html">License</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="../index.html">ReAgent</a>
</nav>
<div class="wy-nav-content">
<div class="rst-content">
<div role="navigation" aria-label="Page navigation">
<ul class="wy-breadcrumbs">
<li><a href="../index.html" class="icon icon-home"></a> &raquo;</li>
<li>reagent.data package</li>
<li class="wy-breadcrumbs-aside">
<a href="../_sources/api/reagent.data.rst.txt" rel="nofollow"> View page source</a>
</li>
</ul>
<hr/>
</div>
<div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
<div itemprop="articleBody">
<section id="reagent-data-package">
<h1>reagent.data package<a class="headerlink" href="#reagent-data-package" title="Permalink to this headline"></a></h1>
<section id="submodules">
<h2>Submodules<a class="headerlink" href="#submodules" title="Permalink to this headline"></a></h2>
</section>
<section id="module-reagent.data.data_fetcher">
<span id="reagent-data-data-fetcher-module"></span><h2>reagent.data.data_fetcher module<a class="headerlink" href="#module-reagent.data.data_fetcher" title="Permalink to this headline"></a></h2>
<dl class="py class">
<dt class="sig sig-object py" id="reagent.data.data_fetcher.DataFetcher">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">reagent.data.data_fetcher.</span></span><span class="sig-name descname"><span class="pre">DataFetcher</span></span><a class="headerlink" href="#reagent.data.data_fetcher.DataFetcher" title="Permalink to this definition"></a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">object</span></code></p>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.data_fetcher.DataFetcher.query_data">
<span class="sig-name descname"><span class="pre">query_data</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">discrete_action</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">bool</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">actions</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">include_possible_actions</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">custom_reward_expression</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">sample_range</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">gamma</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></span><a class="headerlink" href="#reagent.data.data_fetcher.DataFetcher.query_data" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.data_fetcher.DataFetcher.query_data_synthetic_reward">
<span class="sig-name descname"><span class="pre">query_data_synthetic_reward</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">discrete_action_names</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">sample_range</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">max_seq_len</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></span><a class="headerlink" href="#reagent.data.data_fetcher.DataFetcher.query_data_synthetic_reward" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
</dd></dl>
</section>
<section id="module-reagent.data.manual_data_module">
<span id="reagent-data-manual-data-module-module"></span><h2>reagent.data.manual_data_module module<a class="headerlink" href="#module-reagent.data.manual_data_module" title="Permalink to this headline"></a></h2>
<dl class="py class">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">reagent.data.manual_data_module.</span></span><span class="sig-name descname"><span class="pre">ManualDataModule</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="o"><span class="pre">*</span></span><span class="n"><span class="pre">args</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Any</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">kwargs</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Any</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule" title="Permalink to this definition"></a></dt>
<dd><p>Bases: <a class="reference internal" href="#reagent.data.reagent_data_module.ReAgentDataModule" title="reagent.data.reagent_data_module.ReAgentDataModule"><code class="xref py py-class docutils literal notranslate"><span class="pre">reagent.data.reagent_data_module.ReAgentDataModule</span></code></a></p>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.build_batch_preprocessor">
<em class="property"><span class="pre">abstract</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">build_batch_preprocessor</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.preprocessing.html#reagent.preprocessing.batch_preprocessor.BatchPreprocessor" title="reagent.preprocessing.batch_preprocessor.BatchPreprocessor"><span class="pre">reagent.preprocessing.batch_preprocessor.BatchPreprocessor</span></a></span></span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.build_batch_preprocessor" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.get_dataloader">
<span class="sig-name descname"><span class="pre">get_dataloader</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">dataset</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">identity</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">str</span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">'Default'</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.get_dataloader" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.get_normalization_data_map">
<span class="sig-name descname"><span class="pre">get_normalization_data_map</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">keys</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">Dict</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="reagent.core.html#reagent.core.parameters.NormalizationData" title="reagent.core.parameters.NormalizationData"><span class="pre">reagent.core.parameters.NormalizationData</span></a><span class="p"><span class="pre">]</span></span></span></span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.get_normalization_data_map" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py property">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.model_manager">
<em class="property"><span class="pre">property</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">model_manager</span></span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.model_manager" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.prepare_data">
<span class="sig-name descname"><span class="pre">prepare_data</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="o"><span class="pre">*</span></span><span class="n"><span class="pre">args</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">kwargs</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.prepare_data" title="Permalink to this definition"></a></dt>
<dd><p>Use this to download and prepare data.</p>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p>DO NOT set state to the model (use <cite>setup</cite> instead)
since this is NOT called on every GPU in DDP/TPU</p>
</div>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">prepare_data</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="c1"># good</span>
<span class="n">download_data</span><span class="p">()</span>
<span class="n">tokenize</span><span class="p">()</span>
<span class="n">etc</span><span class="p">()</span>
<span class="c1"># bad</span>
<span class="bp">self</span><span class="o">.</span><span class="n">split</span> <span class="o">=</span> <span class="n">data_split</span>
<span class="bp">self</span><span class="o">.</span><span class="n">some_state</span> <span class="o">=</span> <span class="n">some_other_state</span><span class="p">()</span>
</pre></div>
</div>
<p>In DDP prepare_data can be called in two ways (using Trainer(prepare_data_per_node)):</p>
<ol class="arabic simple">
<li><p>Once per node. This is the default and is only called on LOCAL_RANK=0.</p></li>
<li><p>Once in total. Only called on GLOBAL_RANK=0.</p></li>
</ol>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># DEFAULT</span>
<span class="c1"># called once per node on LOCAL_RANK=0 of that node</span>
<span class="n">Trainer</span><span class="p">(</span><span class="n">prepare_data_per_node</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="c1"># call on GLOBAL_RANK=0 (great for shared file systems)</span>
<span class="n">Trainer</span><span class="p">(</span><span class="n">prepare_data_per_node</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span>
</pre></div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Setting <code class="docutils literal notranslate"><span class="pre">prepare_data_per_node</span></code> with the trainer flag is deprecated and will be removed in v1.7.0.
Please set <code class="docutils literal notranslate"><span class="pre">prepare_data_per_node</span></code> in LightningDataModule or LightningModule directly instead.</p>
</div>
<p>This is called before requesting the dataloaders:</p>
<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">model</span><span class="o">.</span><span class="n">prepare_data</span><span class="p">()</span>
<span class="n">initialize_distributed</span><span class="p">()</span>
<span class="n">model</span><span class="o">.</span><span class="n">setup</span><span class="p">(</span><span class="n">stage</span><span class="p">)</span>
<span class="n">model</span><span class="o">.</span><span class="n">train_dataloader</span><span class="p">()</span>
<span class="n">model</span><span class="o">.</span><span class="n">val_dataloader</span><span class="p">()</span>
<span class="n">model</span><span class="o">.</span><span class="n">test_dataloader</span><span class="p">()</span>
</pre></div>
</div>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.query_data">
<em class="property"><span class="pre">abstract</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">query_data</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">sample_range</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span></em>, <em class="sig-param"><span class="n"><span class="pre">reward_options</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.RewardOptions" title="reagent.workflow.types.RewardOptions"><span class="pre">reagent.workflow.types.RewardOptions</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">data_fetcher</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="#reagent.data.data_fetcher.DataFetcher" title="reagent.data.data_fetcher.DataFetcher"><span class="pre">reagent.data.data_fetcher.DataFetcher</span></a></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.query_data" title="Permalink to this definition"></a></dt>
<dd><p>Massage input table into the format expected by the trainer</p>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.run_feature_identification">
<em class="property"><span class="pre">abstract</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">run_feature_identification</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">Dict</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="reagent.core.html#reagent.core.parameters.NormalizationData" title="reagent.core.parameters.NormalizationData"><span class="pre">reagent.core.parameters.NormalizationData</span></a><span class="p"><span class="pre">]</span></span></span></span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.run_feature_identification" title="Permalink to this definition"></a></dt>
<dd><p>Derive preprocessing parameters from data.</p>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.setup">
<span class="sig-name descname"><span class="pre">setup</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">stage</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.setup" title="Permalink to this definition"></a></dt>
<dd><p>Called at the beginning of fit (train + validate), validate, test, and predict. This is a good hook when
you need to build models dynamically or adjust something about them. This hook is called on every process
when using DDP.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><p><strong>stage</strong> – either <code class="docutils literal notranslate"><span class="pre">'fit'</span></code>, <code class="docutils literal notranslate"><span class="pre">'validate'</span></code>, <code class="docutils literal notranslate"><span class="pre">'test'</span></code>, or <code class="docutils literal notranslate"><span class="pre">'predict'</span></code></p>
</dd>
</dl>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">LitModel</span><span class="p">(</span><span class="o">...</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="bp">self</span><span class="o">.</span><span class="n">l1</span> <span class="o">=</span> <span class="kc">None</span>
<span class="k">def</span> <span class="nf">prepare_data</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">download_data</span><span class="p">()</span>
<span class="n">tokenize</span><span class="p">()</span>
<span class="c1"># don&#39;t do this</span>
<span class="bp">self</span><span class="o">.</span><span class="n">something</span> <span class="o">=</span> <span class="k">else</span>
<span class="k">def</span> <span class="nf">setup</span><span class="p">(</span><span class="n">stage</span><span class="p">):</span>
<span class="n">data</span> <span class="o">=</span> <span class="n">Load_data</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">l1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">28</span><span class="p">,</span> <span class="n">data</span><span class="o">.</span><span class="n">num_classes</span><span class="p">)</span>
</pre></div>
</div>
</dd></dl>
<dl class="py property">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.should_generate_eval_dataset">
<em class="property"><span class="pre">abstract</span><span class="w"> </span><span class="pre">property</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">should_generate_eval_dataset</span></span><em class="property"><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="pre">bool</span></em><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.should_generate_eval_dataset" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.test_dataloader">
<span class="sig-name descname"><span class="pre">test_dataloader</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.test_dataloader" title="Permalink to this definition"></a></dt>
<dd><p>Implement one or multiple PyTorch DataLoaders for testing.</p>
<p>The dataloader you return will not be reloaded unless you set
<a href="#id1"><span class="problematic" id="id2">:paramref:`~pytorch_lightning.trainer.Trainer.reload_dataloaders_every_n_epochs`</span></a> to
a postive integer.</p>
<p>For data processing use the following pattern:</p>
<blockquote>
<div><ul class="simple">
<li><p>download in <a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.prepare_data" title="reagent.data.manual_data_module.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p>process and split in <a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.setup" title="reagent.data.manual_data_module.ManualDataModule.setup"><code class="xref py py-meth docutils literal notranslate"><span class="pre">setup()</span></code></a></p></li>
</ul>
</div></blockquote>
<p>However, the above are only necessary for distributed processing.</p>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p>do not assign state in prepare_data</p>
</div>
<ul class="simple">
<li><p><code class="xref py py-meth docutils literal notranslate"><span class="pre">fit()</span></code></p></li>
<li><p>…</p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.prepare_data" title="reagent.data.manual_data_module.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.setup" title="reagent.data.manual_data_module.ManualDataModule.setup"><code class="xref py py-meth docutils literal notranslate"><span class="pre">setup()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.train_dataloader" title="reagent.data.manual_data_module.ManualDataModule.train_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">train_dataloader()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.val_dataloader" title="reagent.data.manual_data_module.ManualDataModule.val_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">val_dataloader()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.test_dataloader" title="reagent.data.manual_data_module.ManualDataModule.test_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">test_dataloader()</span></code></a></p></li>
</ul>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Lightning adds the correct sampler for distributed and arbitrary hardware.
There is no need to set it yourself.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Returns</dt>
<dd class="field-odd"><p>A <code class="xref py py-class docutils literal notranslate"><span class="pre">torch.utils.data.DataLoader</span></code> or a sequence of them specifying testing samples.</p>
</dd>
</dl>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">test_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">transform</span> <span class="o">=</span> <span class="n">transforms</span><span class="o">.</span><span class="n">Compose</span><span class="p">([</span><span class="n">transforms</span><span class="o">.</span><span class="n">ToTensor</span><span class="p">(),</span>
<span class="n">transforms</span><span class="o">.</span><span class="n">Normalize</span><span class="p">((</span><span class="mf">0.5</span><span class="p">,),</span> <span class="p">(</span><span class="mf">1.0</span><span class="p">,))])</span>
<span class="n">dataset</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="n">root</span><span class="o">=</span><span class="s1">&#39;/path/to/mnist/&#39;</span><span class="p">,</span> <span class="n">train</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span> <span class="n">transform</span><span class="o">=</span><span class="n">transform</span><span class="p">,</span>
<span class="n">download</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="n">loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">dataset</span><span class="p">,</span>
<span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span>
<span class="n">shuffle</span><span class="o">=</span><span class="kc">False</span>
<span class="p">)</span>
<span class="k">return</span> <span class="n">loader</span>
<span class="c1"># can also return multiple dataloaders</span>
<span class="k">def</span> <span class="nf">test_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="k">return</span> <span class="p">[</span><span class="n">loader_a</span><span class="p">,</span> <span class="n">loader_b</span><span class="p">,</span> <span class="o">...</span><span class="p">,</span> <span class="n">loader_n</span><span class="p">]</span>
</pre></div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>If you don’t need a test dataset and a <code class="xref py py-meth docutils literal notranslate"><span class="pre">test_step()</span></code>, you don’t need to implement
this method.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>In the case where you return multiple test dataloaders, the <code class="xref py py-meth docutils literal notranslate"><span class="pre">test_step()</span></code>
will have an argument <code class="docutils literal notranslate"><span class="pre">dataloader_idx</span></code> which matches the order here.</p>
</div>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.train_dataloader">
<span class="sig-name descname"><span class="pre">train_dataloader</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.train_dataloader" title="Permalink to this definition"></a></dt>
<dd><p>Implement one or more PyTorch DataLoaders for training.</p>
<dl class="field-list simple">
<dt class="field-odd">Returns</dt>
<dd class="field-odd"><p>A collection of <code class="xref py py-class docutils literal notranslate"><span class="pre">torch.utils.data.DataLoader</span></code> specifying training samples.
In the case of multiple dataloaders, please see this <span class="xref std std-ref">page</span>.</p>
</dd>
</dl>
<p>The dataloader you return will not be reloaded unless you set
<a href="#id3"><span class="problematic" id="id4">:paramref:`~pytorch_lightning.trainer.Trainer.reload_dataloaders_every_n_epochs`</span></a> to
a positive integer.</p>
<p>For data processing use the following pattern:</p>
<blockquote>
<div><ul class="simple">
<li><p>download in <a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.prepare_data" title="reagent.data.manual_data_module.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p>process and split in <a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.setup" title="reagent.data.manual_data_module.ManualDataModule.setup"><code class="xref py py-meth docutils literal notranslate"><span class="pre">setup()</span></code></a></p></li>
</ul>
</div></blockquote>
<p>However, the above are only necessary for distributed processing.</p>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p>do not assign state in prepare_data</p>
</div>
<ul class="simple">
<li><p><code class="xref py py-meth docutils literal notranslate"><span class="pre">fit()</span></code></p></li>
<li><p>…</p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.prepare_data" title="reagent.data.manual_data_module.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.setup" title="reagent.data.manual_data_module.ManualDataModule.setup"><code class="xref py py-meth docutils literal notranslate"><span class="pre">setup()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.train_dataloader" title="reagent.data.manual_data_module.ManualDataModule.train_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">train_dataloader()</span></code></a></p></li>
</ul>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Lightning adds the correct sampler for distributed and arbitrary hardware.
There is no need to set it yourself.</p>
</div>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># single dataloader</span>
<span class="k">def</span> <span class="nf">train_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">transform</span> <span class="o">=</span> <span class="n">transforms</span><span class="o">.</span><span class="n">Compose</span><span class="p">([</span><span class="n">transforms</span><span class="o">.</span><span class="n">ToTensor</span><span class="p">(),</span>
<span class="n">transforms</span><span class="o">.</span><span class="n">Normalize</span><span class="p">((</span><span class="mf">0.5</span><span class="p">,),</span> <span class="p">(</span><span class="mf">1.0</span><span class="p">,))])</span>
<span class="n">dataset</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="n">root</span><span class="o">=</span><span class="s1">&#39;/path/to/mnist/&#39;</span><span class="p">,</span> <span class="n">train</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="n">transform</span><span class="o">=</span><span class="n">transform</span><span class="p">,</span>
<span class="n">download</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="n">loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">dataset</span><span class="p">,</span>
<span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span>
<span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="k">return</span> <span class="n">loader</span>
<span class="c1"># multiple dataloaders, return as list</span>
<span class="k">def</span> <span class="nf">train_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">mnist</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">cifar</span> <span class="o">=</span> <span class="n">CIFAR</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">mnist_loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">mnist</span><span class="p">,</span> <span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="n">cifar_loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">cifar</span><span class="p">,</span> <span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="c1"># each batch will be a list of tensors: [batch_mnist, batch_cifar]</span>
<span class="k">return</span> <span class="p">[</span><span class="n">mnist_loader</span><span class="p">,</span> <span class="n">cifar_loader</span><span class="p">]</span>
<span class="c1"># multiple dataloader, return as dict</span>
<span class="k">def</span> <span class="nf">train_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">mnist</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">cifar</span> <span class="o">=</span> <span class="n">CIFAR</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">mnist_loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">mnist</span><span class="p">,</span> <span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="n">cifar_loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">cifar</span><span class="p">,</span> <span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="c1"># each batch will be a dict of tensors: {&#39;mnist&#39;: batch_mnist, &#39;cifar&#39;: batch_cifar}</span>
<span class="k">return</span> <span class="p">{</span><span class="s1">&#39;mnist&#39;</span><span class="p">:</span> <span class="n">mnist_loader</span><span class="p">,</span> <span class="s1">&#39;cifar&#39;</span><span class="p">:</span> <span class="n">cifar_loader</span><span class="p">}</span>
</pre></div>
</div>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.ManualDataModule.val_dataloader">
<span class="sig-name descname"><span class="pre">val_dataloader</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.manual_data_module.ManualDataModule.val_dataloader" title="Permalink to this definition"></a></dt>
<dd><p>Implement one or multiple PyTorch DataLoaders for validation.</p>
<p>The dataloader you return will not be reloaded unless you set
<a href="#id5"><span class="problematic" id="id6">:paramref:`~pytorch_lightning.trainer.Trainer.reload_dataloaders_every_n_epochs`</span></a> to
a positive integer.</p>
<p>It’s recommended that all data downloads and preparation happen in <a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.prepare_data" title="reagent.data.manual_data_module.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a>.</p>
<ul class="simple">
<li><p><code class="xref py py-meth docutils literal notranslate"><span class="pre">fit()</span></code></p></li>
<li><p>…</p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.prepare_data" title="reagent.data.manual_data_module.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.train_dataloader" title="reagent.data.manual_data_module.ManualDataModule.train_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">train_dataloader()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.val_dataloader" title="reagent.data.manual_data_module.ManualDataModule.val_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">val_dataloader()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.manual_data_module.ManualDataModule.test_dataloader" title="reagent.data.manual_data_module.ManualDataModule.test_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">test_dataloader()</span></code></a></p></li>
</ul>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Lightning adds the correct sampler for distributed and arbitrary hardware
There is no need to set it yourself.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Returns</dt>
<dd class="field-odd"><p>A <code class="xref py py-class docutils literal notranslate"><span class="pre">torch.utils.data.DataLoader</span></code> or a sequence of them specifying validation samples.</p>
</dd>
</dl>
<p>Examples:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">val_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">transform</span> <span class="o">=</span> <span class="n">transforms</span><span class="o">.</span><span class="n">Compose</span><span class="p">([</span><span class="n">transforms</span><span class="o">.</span><span class="n">ToTensor</span><span class="p">(),</span>
<span class="n">transforms</span><span class="o">.</span><span class="n">Normalize</span><span class="p">((</span><span class="mf">0.5</span><span class="p">,),</span> <span class="p">(</span><span class="mf">1.0</span><span class="p">,))])</span>
<span class="n">dataset</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="n">root</span><span class="o">=</span><span class="s1">&#39;/path/to/mnist/&#39;</span><span class="p">,</span> <span class="n">train</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span>
<span class="n">transform</span><span class="o">=</span><span class="n">transform</span><span class="p">,</span> <span class="n">download</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="n">loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">dataset</span><span class="p">,</span>
<span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span>
<span class="n">shuffle</span><span class="o">=</span><span class="kc">False</span>
<span class="p">)</span>
<span class="k">return</span> <span class="n">loader</span>
<span class="c1"># can also return multiple dataloaders</span>
<span class="k">def</span> <span class="nf">val_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="k">return</span> <span class="p">[</span><span class="n">loader_a</span><span class="p">,</span> <span class="n">loader_b</span><span class="p">,</span> <span class="o">...</span><span class="p">,</span> <span class="n">loader_n</span><span class="p">]</span>
</pre></div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>If you don’t need a validation dataset and a <code class="xref py py-meth docutils literal notranslate"><span class="pre">validation_step()</span></code>, you don’t need to
implement this method.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>In the case where you return multiple validation dataloaders, the <code class="xref py py-meth docutils literal notranslate"><span class="pre">validation_step()</span></code>
will have an argument <code class="docutils literal notranslate"><span class="pre">dataloader_idx</span></code> which matches the order here.</p>
</div>
</dd></dl>
</dd></dl>
<dl class="py class">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.TrainEvalSampleRanges">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">reagent.data.manual_data_module.</span></span><span class="sig-name descname"><span class="pre">TrainEvalSampleRanges</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">train_sample_range</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">eval_sample_range</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.manual_data_module.TrainEvalSampleRanges" title="Permalink to this definition"></a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">NamedTuple</span></code></p>
<dl class="py attribute">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.TrainEvalSampleRanges.eval_sample_range">
<span class="sig-name descname"><span class="pre">eval_sample_range</span></span><em class="property"><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span></em><a class="headerlink" href="#reagent.data.manual_data_module.TrainEvalSampleRanges.eval_sample_range" title="Permalink to this definition"></a></dt>
<dd><p>Alias for field number 1</p>
</dd></dl>
<dl class="py attribute">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.TrainEvalSampleRanges.train_sample_range">
<span class="sig-name descname"><span class="pre">train_sample_range</span></span><em class="property"><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span></em><a class="headerlink" href="#reagent.data.manual_data_module.TrainEvalSampleRanges.train_sample_range" title="Permalink to this definition"></a></dt>
<dd><p>Alias for field number 0</p>
</dd></dl>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.collate_and_preprocess">
<span class="sig-prename descclassname"><span class="pre">reagent.data.manual_data_module.</span></span><span class="sig-name descname"><span class="pre">collate_and_preprocess</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">batch_preprocessor</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.preprocessing.html#reagent.preprocessing.batch_preprocessor.BatchPreprocessor" title="reagent.preprocessing.batch_preprocessor.BatchPreprocessor"><span class="pre">reagent.preprocessing.batch_preprocessor.BatchPreprocessor</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">use_gpu</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">bool</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.manual_data_module.collate_and_preprocess" title="Permalink to this definition"></a></dt>
<dd><p>Helper for Petastorm’s DataLoader to preprocess.
TODO(kaiwenw): parallelize preprocessing by using transform of Petastorm reader
Should pin memory and preprocess in reader and convert to gpu in collate_fn.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.manual_data_module.get_sample_range">
<span class="sig-prename descclassname"><span class="pre">reagent.data.manual_data_module.</span></span><span class="sig-name descname"><span class="pre">get_sample_range</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">calc_cpe_in_training</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">bool</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="#reagent.data.manual_data_module.TrainEvalSampleRanges" title="reagent.data.manual_data_module.TrainEvalSampleRanges"><span class="pre">reagent.data.manual_data_module.TrainEvalSampleRanges</span></a></span></span><a class="headerlink" href="#reagent.data.manual_data_module.get_sample_range" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
</section>
<section id="module-reagent.data.oss_data_fetcher">
<span id="reagent-data-oss-data-fetcher-module"></span><h2>reagent.data.oss_data_fetcher module<a class="headerlink" href="#module-reagent.data.oss_data_fetcher" title="Permalink to this headline"></a></h2>
<dl class="py class">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.OssDataFetcher">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">OssDataFetcher</span></span><a class="headerlink" href="#reagent.data.oss_data_fetcher.OssDataFetcher" title="Permalink to this definition"></a></dt>
<dd><p>Bases: <a class="reference internal" href="#reagent.data.data_fetcher.DataFetcher" title="reagent.data.data_fetcher.DataFetcher"><code class="xref py py-class docutils literal notranslate"><span class="pre">reagent.data.data_fetcher.DataFetcher</span></code></a></p>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.OssDataFetcher.query_data">
<span class="sig-name descname"><span class="pre">query_data</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">discrete_action</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">bool</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">actions</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">include_possible_actions</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">custom_reward_expression</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">sample_range</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">gamma</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></span><a class="headerlink" href="#reagent.data.oss_data_fetcher.OssDataFetcher.query_data" title="Permalink to this definition"></a></dt>
<dd><p>Perform reward calculation, hashing mdp + subsampling and
other preprocessing such as sparse2dense.</p>
</dd></dl>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.calc_custom_reward">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">calc_custom_reward</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">custom_reward_expression</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">str</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.calc_custom_reward" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.calc_reward_multi_steps">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">calc_reward_multi_steps</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">int</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">gamma</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">float</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.calc_reward_multi_steps" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.discrete_action_preprocessing">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">discrete_action_preprocessing</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">actions</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.discrete_action_preprocessing" title="Permalink to this definition"></a></dt>
<dd><dl>
<dt>Inputted actions and possible_actions are strings, which isn’t supported</dt><dd><p>for PyTorch Tensors. Here, we represent them with LongType.
(a) action and next_action are strings, so simply return their position</p>
<blockquote>
<div><p>in the action_space (as given by argument actions).</p>
</div></blockquote>
<ol class="loweralpha simple" start="2">
<li><p>possible_actions and possible_next_actions are list of strs, so
return an existence bitvector of length len(actions), where ith
index is true iff actions[i] was in the list.</p></li>
</ol>
</dd>
</dl>
<p>By-product: output not_terminal from preprocessing actions.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.get_distinct_keys">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">get_distinct_keys</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">col_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">is_col_arr_map</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.get_distinct_keys" title="Permalink to this definition"></a></dt>
<dd><p>Return list of distinct keys.
Set is_col_arr_map to be true if column is an array of Maps.
Otherwise, assume column is a Map.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.hash_mdp_id_and_subsample">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">hash_mdp_id_and_subsample</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">sample_range</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.hash_mdp_id_and_subsample" title="Permalink to this definition"></a></dt>
<dd><p>Since mdp_id is a string but Pytorch Tensors do not store strings,
we hash them with crc32, which is treated as a cryptographic hash
(with range [0, MAX_UINT32-1]). We also perform an optional subsampling
based on this hash value.
NOTE: we’re assuming no collisions in this hash! Otherwise, two mdp_ids
can be indistinguishable after the hash.
TODO: change this to a deterministic subsample.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.infer_action_names">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">infer_action_names</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.infer_action_names" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.infer_metrics_names">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">infer_metrics_names</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.infer_metrics_names" title="Permalink to this definition"></a></dt>
<dd><p>Infer possible metrics names.
Assume in multi-step case, metrics is an array of maps.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.infer_states_names">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">infer_states_names</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.infer_states_names" title="Permalink to this definition"></a></dt>
<dd><p>Infer possible state names from states and next state features.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.make_existence_bitvector_udf">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">make_existence_bitvector_udf</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">arr</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.make_existence_bitvector_udf" title="Permalink to this definition"></a></dt>
<dd><p>one-hot encode elements of target depending on their existence in arr.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.make_get_step_udf">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">make_get_step_udf</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.make_get_step_udf" title="Permalink to this definition"></a></dt>
<dd><p>Get step count by taking length of next_states_features array.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.make_next_udf">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">make_next_udf</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span></em>, <em class="sig-param"><span class="n"><span class="pre">return_type</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.make_next_udf" title="Permalink to this definition"></a></dt>
<dd><p>Generic udf to get next (after multi_steps) item, provided item type.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.make_sparse2dense">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">make_sparse2dense</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">col_name</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">str</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">possible_keys</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">List</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.make_sparse2dense" title="Permalink to this definition"></a></dt>
<dd><p>Given a list of possible keys, convert sparse map to dense array.
In our example, both value_type is assumed to be a float.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.make_where_udf">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">make_where_udf</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">arr</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.make_where_udf" title="Permalink to this definition"></a></dt>
<dd><p>Return index of item in arr, and len(arr) if not found.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.misc_column_preprocessing">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">misc_column_preprocessing</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.misc_column_preprocessing" title="Permalink to this definition"></a></dt>
<dd><p>Miscellaneous columns are step, time_diff, sequence_number, not_terminal.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.parametric_action_preprocessing">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">parametric_action_preprocessing</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">actions</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">include_possible_actions</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">bool</span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">True</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.parametric_action_preprocessing" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.rand_string">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">rand_string</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">length</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.rand_string" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.select_relevant_columns">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">select_relevant_columns</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">discrete_action</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">bool</span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">include_possible_actions</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">bool</span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">True</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.select_relevant_columns" title="Permalink to this definition"></a></dt>
<dd><p>Select all the relevant columns and perform type conversions.</p>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.set_reward_col_as_reward">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">set_reward_col_as_reward</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">custom_reward_expression</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">gamma</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.set_reward_col_as_reward" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.state_and_metrics_sparse2dense">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">state_and_metrics_sparse2dense</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">states</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span></em>, <em class="sig-param"><span class="n"><span class="pre">metrics</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.oss_data_fetcher.state_and_metrics_sparse2dense" title="Permalink to this definition"></a></dt>
<dd><p>Sparse-to-dense preprocessing of Map columns, which are states and metrics.
For each column of type Map, w/ name X, output two columns.</p>
<blockquote>
<div><p>Map values are assumed to be scalar. This process is called sparse-to-dense.
X = {“state_features”, “next_state_features”, “metrics”}.
(a) Replace column X with a dense repesentation of the inputted (sparse) map.</p>
<blockquote>
<div><p>Dense representation is to concatenate map values into a list.</p>
</div></blockquote>
<ol class="loweralpha simple" start="2">
<li><p>Create new column X_presence, which is a list of same length as (a) and
the ith entry is 1 iff the key was present in the original map.</p></li>
</ol>
</div></blockquote>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.oss_data_fetcher.upload_as_parquet">
<span class="sig-prename descclassname"><span class="pre">reagent.data.oss_data_fetcher.</span></span><span class="sig-name descname"><span class="pre">upload_as_parquet</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">df</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></span><a class="headerlink" href="#reagent.data.oss_data_fetcher.upload_as_parquet" title="Permalink to this definition"></a></dt>
<dd><p>Generate a random parquet. Fails if cannot generate a non-existent name.</p>
</dd></dl>
</section>
<section id="module-reagent.data.reagent_data_module">
<span id="reagent-data-reagent-data-module-module"></span><h2>reagent.data.reagent_data_module module<a class="headerlink" href="#module-reagent.data.reagent_data_module" title="Permalink to this headline"></a></h2>
<dl class="py class">
<dt class="sig sig-object py" id="reagent.data.reagent_data_module.ReAgentDataModule">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">reagent.data.reagent_data_module.</span></span><span class="sig-name descname"><span class="pre">ReAgentDataModule</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="o"><span class="pre">*</span></span><span class="n"><span class="pre">args</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Any</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">kwargs</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Any</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.reagent_data_module.ReAgentDataModule" title="Permalink to this definition"></a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">pytorch_lightning.core.datamodule.LightningDataModule</span></code></p>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.reagent_data_module.ReAgentDataModule.get_normalization_data_map">
<em class="property"><span class="pre">abstract</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">get_normalization_data_map</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">keys</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">Dict</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="reagent.core.html#reagent.core.parameters.NormalizationData" title="reagent.core.parameters.NormalizationData"><span class="pre">reagent.core.parameters.NormalizationData</span></a><span class="p"><span class="pre">]</span></span></span></span><a class="headerlink" href="#reagent.data.reagent_data_module.ReAgentDataModule.get_normalization_data_map" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
</dd></dl>
</section>
<section id="module-reagent.data.spark_utils">
<span id="reagent-data-spark-utils-module"></span><h2>reagent.data.spark_utils module<a class="headerlink" href="#module-reagent.data.spark_utils" title="Permalink to this headline"></a></h2>
<dl class="py data">
<dt class="sig sig-object py" id="reagent.data.spark_utils.SPARK_JAR_FROM_ROOT_DIR">
<span class="sig-prename descclassname"><span class="pre">reagent.data.spark_utils.</span></span><span class="sig-name descname"><span class="pre">SPARK_JAR_FROM_ROOT_DIR</span></span><em class="property"><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="pre">'preprocessing/target/rl-preprocessing-1.1.jar'</span></em><a class="headerlink" href="#reagent.data.spark_utils.SPARK_JAR_FROM_ROOT_DIR" title="Permalink to this definition"></a></dt>
<dd><p>SPARK_JAR is abspath to the above jar file.</p>
<p>Assume file structure
ReAgent/</p>
<blockquote>
<div><p>preprocessing/…
reagent/…</p>
</div></blockquote>
</dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.spark_utils.call_spark_class">
<span class="sig-prename descclassname"><span class="pre">reagent.data.spark_utils.</span></span><span class="sig-name descname"><span class="pre">call_spark_class</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">spark</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">class_name</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">str</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">args</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">str</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.spark_utils.call_spark_class" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.spark_utils.get_spark_session">
<span class="sig-prename descclassname"><span class="pre">reagent.data.spark_utils.</span></span><span class="sig-name descname"><span class="pre">get_spark_session</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">config</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">Dict</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">{'spark.app.name':</span> <span class="pre">'ReAgent',</span> <span class="pre">'spark.driver.extraClassPath':</span> <span class="pre">'/Users/czxttkl/github/ReAgent/preprocessing/target/rl-preprocessing-1.1.jar',</span> <span class="pre">'spark.driver.host':</span> <span class="pre">'127.0.0.1',</span> <span class="pre">'spark.master':</span> <span class="pre">'local[*]',</span> <span class="pre">'spark.sql.catalogImplementation':</span> <span class="pre">'hive',</span> <span class="pre">'spark.sql.execution.arrow.enabled':</span> <span class="pre">'true',</span> <span class="pre">'spark.sql.session.timeZone':</span> <span class="pre">'UTC',</span> <span class="pre">'spark.sql.shuffle.partitions':</span> <span class="pre">'12',</span> <span class="pre">'spark.sql.warehouse.dir':</span> <span class="pre">'/Users/czxttkl/github/ReAgent/docs/spark-warehouse'}</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.spark_utils.get_spark_session" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py function">
<dt class="sig sig-object py" id="reagent.data.spark_utils.get_table_url">
<span class="sig-prename descclassname"><span class="pre">reagent.data.spark_utils.</span></span><span class="sig-name descname"><span class="pre">get_table_url</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">table_name</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">str</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">str</span></span></span><a class="headerlink" href="#reagent.data.spark_utils.get_table_url" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
</section>
<section id="module-reagent.data">
<span id="module-contents"></span><h2>Module contents<a class="headerlink" href="#module-reagent.data" title="Permalink to this headline"></a></h2>
<dl class="py class">
<dt class="sig sig-object py" id="reagent.data.DataFetcher">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">reagent.data.</span></span><span class="sig-name descname"><span class="pre">DataFetcher</span></span><a class="headerlink" href="#reagent.data.DataFetcher" title="Permalink to this definition"></a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">object</span></code></p>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.DataFetcher.query_data">
<span class="sig-name descname"><span class="pre">query_data</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">discrete_action</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">bool</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">actions</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">include_possible_actions</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">custom_reward_expression</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">sample_range</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">multi_steps</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">gamma</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></span><a class="headerlink" href="#reagent.data.DataFetcher.query_data" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.DataFetcher.query_data_synthetic_reward">
<span class="sig-name descname"><span class="pre">query_data_synthetic_reward</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">discrete_action_names</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">sample_range</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">max_seq_len</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></span><a class="headerlink" href="#reagent.data.DataFetcher.query_data_synthetic_reward" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
</dd></dl>
<dl class="py class">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">reagent.data.</span></span><span class="sig-name descname"><span class="pre">ManualDataModule</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="o"><span class="pre">*</span></span><span class="n"><span class="pre">args</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Any</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">kwargs</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Any</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.ManualDataModule" title="Permalink to this definition"></a></dt>
<dd><p>Bases: <a class="reference internal" href="#reagent.data.reagent_data_module.ReAgentDataModule" title="reagent.data.reagent_data_module.ReAgentDataModule"><code class="xref py py-class docutils literal notranslate"><span class="pre">reagent.data.reagent_data_module.ReAgentDataModule</span></code></a></p>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.build_batch_preprocessor">
<em class="property"><span class="pre">abstract</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">build_batch_preprocessor</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.preprocessing.html#reagent.preprocessing.batch_preprocessor.BatchPreprocessor" title="reagent.preprocessing.batch_preprocessor.BatchPreprocessor"><span class="pre">reagent.preprocessing.batch_preprocessor.BatchPreprocessor</span></a></span></span><a class="headerlink" href="#reagent.data.ManualDataModule.build_batch_preprocessor" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.get_dataloader">
<span class="sig-name descname"><span class="pre">get_dataloader</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">dataset</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">identity</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">str</span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">'Default'</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.ManualDataModule.get_dataloader" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.get_normalization_data_map">
<span class="sig-name descname"><span class="pre">get_normalization_data_map</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">keys</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">Dict</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="reagent.core.html#reagent.core.parameters.NormalizationData" title="reagent.core.parameters.NormalizationData"><span class="pre">reagent.core.parameters.NormalizationData</span></a><span class="p"><span class="pre">]</span></span></span></span><a class="headerlink" href="#reagent.data.ManualDataModule.get_normalization_data_map" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py property">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.model_manager">
<em class="property"><span class="pre">property</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">model_manager</span></span><a class="headerlink" href="#reagent.data.ManualDataModule.model_manager" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.prepare_data">
<span class="sig-name descname"><span class="pre">prepare_data</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="o"><span class="pre">*</span></span><span class="n"><span class="pre">args</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">kwargs</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.ManualDataModule.prepare_data" title="Permalink to this definition"></a></dt>
<dd><p>Use this to download and prepare data.</p>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p>DO NOT set state to the model (use <cite>setup</cite> instead)
since this is NOT called on every GPU in DDP/TPU</p>
</div>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">prepare_data</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="c1"># good</span>
<span class="n">download_data</span><span class="p">()</span>
<span class="n">tokenize</span><span class="p">()</span>
<span class="n">etc</span><span class="p">()</span>
<span class="c1"># bad</span>
<span class="bp">self</span><span class="o">.</span><span class="n">split</span> <span class="o">=</span> <span class="n">data_split</span>
<span class="bp">self</span><span class="o">.</span><span class="n">some_state</span> <span class="o">=</span> <span class="n">some_other_state</span><span class="p">()</span>
</pre></div>
</div>
<p>In DDP prepare_data can be called in two ways (using Trainer(prepare_data_per_node)):</p>
<ol class="arabic simple">
<li><p>Once per node. This is the default and is only called on LOCAL_RANK=0.</p></li>
<li><p>Once in total. Only called on GLOBAL_RANK=0.</p></li>
</ol>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># DEFAULT</span>
<span class="c1"># called once per node on LOCAL_RANK=0 of that node</span>
<span class="n">Trainer</span><span class="p">(</span><span class="n">prepare_data_per_node</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="c1"># call on GLOBAL_RANK=0 (great for shared file systems)</span>
<span class="n">Trainer</span><span class="p">(</span><span class="n">prepare_data_per_node</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span>
</pre></div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Setting <code class="docutils literal notranslate"><span class="pre">prepare_data_per_node</span></code> with the trainer flag is deprecated and will be removed in v1.7.0.
Please set <code class="docutils literal notranslate"><span class="pre">prepare_data_per_node</span></code> in LightningDataModule or LightningModule directly instead.</p>
</div>
<p>This is called before requesting the dataloaders:</p>
<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">model</span><span class="o">.</span><span class="n">prepare_data</span><span class="p">()</span>
<span class="n">initialize_distributed</span><span class="p">()</span>
<span class="n">model</span><span class="o">.</span><span class="n">setup</span><span class="p">(</span><span class="n">stage</span><span class="p">)</span>
<span class="n">model</span><span class="o">.</span><span class="n">train_dataloader</span><span class="p">()</span>
<span class="n">model</span><span class="o">.</span><span class="n">val_dataloader</span><span class="p">()</span>
<span class="n">model</span><span class="o">.</span><span class="n">test_dataloader</span><span class="p">()</span>
</pre></div>
</div>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.query_data">
<em class="property"><span class="pre">abstract</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">query_data</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">sample_range</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">Tuple</span><span class="p"><span class="pre">[</span></span><span class="pre">float</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="pre">float</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span></em>, <em class="sig-param"><span class="n"><span class="pre">reward_options</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.RewardOptions" title="reagent.workflow.types.RewardOptions"><span class="pre">reagent.workflow.types.RewardOptions</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">data_fetcher</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="#reagent.data.data_fetcher.DataFetcher" title="reagent.data.data_fetcher.DataFetcher"><span class="pre">reagent.data.data_fetcher.DataFetcher</span></a></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.Dataset" title="reagent.workflow.types.Dataset"><span class="pre">reagent.workflow.types.Dataset</span></a></span></span><a class="headerlink" href="#reagent.data.ManualDataModule.query_data" title="Permalink to this definition"></a></dt>
<dd><p>Massage input table into the format expected by the trainer</p>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.run_feature_identification">
<em class="property"><span class="pre">abstract</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">run_feature_identification</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_table_spec</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><a class="reference internal" href="reagent.workflow.html#reagent.workflow.types.TableSpec" title="reagent.workflow.types.TableSpec"><span class="pre">reagent.workflow.types.TableSpec</span></a></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">Dict</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="reagent.core.html#reagent.core.parameters.NormalizationData" title="reagent.core.parameters.NormalizationData"><span class="pre">reagent.core.parameters.NormalizationData</span></a><span class="p"><span class="pre">]</span></span></span></span><a class="headerlink" href="#reagent.data.ManualDataModule.run_feature_identification" title="Permalink to this definition"></a></dt>
<dd><p>Derive preprocessing parameters from data.</p>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.setup">
<span class="sig-name descname"><span class="pre">setup</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">stage</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.ManualDataModule.setup" title="Permalink to this definition"></a></dt>
<dd><p>Called at the beginning of fit (train + validate), validate, test, and predict. This is a good hook when
you need to build models dynamically or adjust something about them. This hook is called on every process
when using DDP.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><p><strong>stage</strong> – either <code class="docutils literal notranslate"><span class="pre">'fit'</span></code>, <code class="docutils literal notranslate"><span class="pre">'validate'</span></code>, <code class="docutils literal notranslate"><span class="pre">'test'</span></code>, or <code class="docutils literal notranslate"><span class="pre">'predict'</span></code></p>
</dd>
</dl>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">LitModel</span><span class="p">(</span><span class="o">...</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="bp">self</span><span class="o">.</span><span class="n">l1</span> <span class="o">=</span> <span class="kc">None</span>
<span class="k">def</span> <span class="nf">prepare_data</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">download_data</span><span class="p">()</span>
<span class="n">tokenize</span><span class="p">()</span>
<span class="c1"># don&#39;t do this</span>
<span class="bp">self</span><span class="o">.</span><span class="n">something</span> <span class="o">=</span> <span class="k">else</span>
<span class="k">def</span> <span class="nf">setup</span><span class="p">(</span><span class="n">stage</span><span class="p">):</span>
<span class="n">data</span> <span class="o">=</span> <span class="n">Load_data</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">l1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">28</span><span class="p">,</span> <span class="n">data</span><span class="o">.</span><span class="n">num_classes</span><span class="p">)</span>
</pre></div>
</div>
</dd></dl>
<dl class="py property">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.should_generate_eval_dataset">
<em class="property"><span class="pre">abstract</span><span class="w"> </span><span class="pre">property</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">should_generate_eval_dataset</span></span><em class="property"><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="pre">bool</span></em><a class="headerlink" href="#reagent.data.ManualDataModule.should_generate_eval_dataset" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.test_dataloader">
<span class="sig-name descname"><span class="pre">test_dataloader</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.ManualDataModule.test_dataloader" title="Permalink to this definition"></a></dt>
<dd><p>Implement one or multiple PyTorch DataLoaders for testing.</p>
<p>The dataloader you return will not be reloaded unless you set
<a href="#id7"><span class="problematic" id="id8">:paramref:`~pytorch_lightning.trainer.Trainer.reload_dataloaders_every_n_epochs`</span></a> to
a postive integer.</p>
<p>For data processing use the following pattern:</p>
<blockquote>
<div><ul class="simple">
<li><p>download in <a class="reference internal" href="#reagent.data.ManualDataModule.prepare_data" title="reagent.data.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p>process and split in <a class="reference internal" href="#reagent.data.ManualDataModule.setup" title="reagent.data.ManualDataModule.setup"><code class="xref py py-meth docutils literal notranslate"><span class="pre">setup()</span></code></a></p></li>
</ul>
</div></blockquote>
<p>However, the above are only necessary for distributed processing.</p>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p>do not assign state in prepare_data</p>
</div>
<ul class="simple">
<li><p><code class="xref py py-meth docutils literal notranslate"><span class="pre">fit()</span></code></p></li>
<li><p>…</p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.prepare_data" title="reagent.data.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.setup" title="reagent.data.ManualDataModule.setup"><code class="xref py py-meth docutils literal notranslate"><span class="pre">setup()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.train_dataloader" title="reagent.data.ManualDataModule.train_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">train_dataloader()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.val_dataloader" title="reagent.data.ManualDataModule.val_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">val_dataloader()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.test_dataloader" title="reagent.data.ManualDataModule.test_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">test_dataloader()</span></code></a></p></li>
</ul>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Lightning adds the correct sampler for distributed and arbitrary hardware.
There is no need to set it yourself.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Returns</dt>
<dd class="field-odd"><p>A <code class="xref py py-class docutils literal notranslate"><span class="pre">torch.utils.data.DataLoader</span></code> or a sequence of them specifying testing samples.</p>
</dd>
</dl>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">test_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">transform</span> <span class="o">=</span> <span class="n">transforms</span><span class="o">.</span><span class="n">Compose</span><span class="p">([</span><span class="n">transforms</span><span class="o">.</span><span class="n">ToTensor</span><span class="p">(),</span>
<span class="n">transforms</span><span class="o">.</span><span class="n">Normalize</span><span class="p">((</span><span class="mf">0.5</span><span class="p">,),</span> <span class="p">(</span><span class="mf">1.0</span><span class="p">,))])</span>
<span class="n">dataset</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="n">root</span><span class="o">=</span><span class="s1">&#39;/path/to/mnist/&#39;</span><span class="p">,</span> <span class="n">train</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span> <span class="n">transform</span><span class="o">=</span><span class="n">transform</span><span class="p">,</span>
<span class="n">download</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="n">loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">dataset</span><span class="p">,</span>
<span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span>
<span class="n">shuffle</span><span class="o">=</span><span class="kc">False</span>
<span class="p">)</span>
<span class="k">return</span> <span class="n">loader</span>
<span class="c1"># can also return multiple dataloaders</span>
<span class="k">def</span> <span class="nf">test_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="k">return</span> <span class="p">[</span><span class="n">loader_a</span><span class="p">,</span> <span class="n">loader_b</span><span class="p">,</span> <span class="o">...</span><span class="p">,</span> <span class="n">loader_n</span><span class="p">]</span>
</pre></div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>If you don’t need a test dataset and a <code class="xref py py-meth docutils literal notranslate"><span class="pre">test_step()</span></code>, you don’t need to implement
this method.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>In the case where you return multiple test dataloaders, the <code class="xref py py-meth docutils literal notranslate"><span class="pre">test_step()</span></code>
will have an argument <code class="docutils literal notranslate"><span class="pre">dataloader_idx</span></code> which matches the order here.</p>
</div>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.train_dataloader">
<span class="sig-name descname"><span class="pre">train_dataloader</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.ManualDataModule.train_dataloader" title="Permalink to this definition"></a></dt>
<dd><p>Implement one or more PyTorch DataLoaders for training.</p>
<dl class="field-list simple">
<dt class="field-odd">Returns</dt>
<dd class="field-odd"><p>A collection of <code class="xref py py-class docutils literal notranslate"><span class="pre">torch.utils.data.DataLoader</span></code> specifying training samples.
In the case of multiple dataloaders, please see this <span class="xref std std-ref">page</span>.</p>
</dd>
</dl>
<p>The dataloader you return will not be reloaded unless you set
<a href="#id9"><span class="problematic" id="id10">:paramref:`~pytorch_lightning.trainer.Trainer.reload_dataloaders_every_n_epochs`</span></a> to
a positive integer.</p>
<p>For data processing use the following pattern:</p>
<blockquote>
<div><ul class="simple">
<li><p>download in <a class="reference internal" href="#reagent.data.ManualDataModule.prepare_data" title="reagent.data.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p>process and split in <a class="reference internal" href="#reagent.data.ManualDataModule.setup" title="reagent.data.ManualDataModule.setup"><code class="xref py py-meth docutils literal notranslate"><span class="pre">setup()</span></code></a></p></li>
</ul>
</div></blockquote>
<p>However, the above are only necessary for distributed processing.</p>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p>do not assign state in prepare_data</p>
</div>
<ul class="simple">
<li><p><code class="xref py py-meth docutils literal notranslate"><span class="pre">fit()</span></code></p></li>
<li><p>…</p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.prepare_data" title="reagent.data.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.setup" title="reagent.data.ManualDataModule.setup"><code class="xref py py-meth docutils literal notranslate"><span class="pre">setup()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.train_dataloader" title="reagent.data.ManualDataModule.train_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">train_dataloader()</span></code></a></p></li>
</ul>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Lightning adds the correct sampler for distributed and arbitrary hardware.
There is no need to set it yourself.</p>
</div>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># single dataloader</span>
<span class="k">def</span> <span class="nf">train_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">transform</span> <span class="o">=</span> <span class="n">transforms</span><span class="o">.</span><span class="n">Compose</span><span class="p">([</span><span class="n">transforms</span><span class="o">.</span><span class="n">ToTensor</span><span class="p">(),</span>
<span class="n">transforms</span><span class="o">.</span><span class="n">Normalize</span><span class="p">((</span><span class="mf">0.5</span><span class="p">,),</span> <span class="p">(</span><span class="mf">1.0</span><span class="p">,))])</span>
<span class="n">dataset</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="n">root</span><span class="o">=</span><span class="s1">&#39;/path/to/mnist/&#39;</span><span class="p">,</span> <span class="n">train</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="n">transform</span><span class="o">=</span><span class="n">transform</span><span class="p">,</span>
<span class="n">download</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="n">loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">dataset</span><span class="p">,</span>
<span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span>
<span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="k">return</span> <span class="n">loader</span>
<span class="c1"># multiple dataloaders, return as list</span>
<span class="k">def</span> <span class="nf">train_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">mnist</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">cifar</span> <span class="o">=</span> <span class="n">CIFAR</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">mnist_loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">mnist</span><span class="p">,</span> <span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="n">cifar_loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">cifar</span><span class="p">,</span> <span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="c1"># each batch will be a list of tensors: [batch_mnist, batch_cifar]</span>
<span class="k">return</span> <span class="p">[</span><span class="n">mnist_loader</span><span class="p">,</span> <span class="n">cifar_loader</span><span class="p">]</span>
<span class="c1"># multiple dataloader, return as dict</span>
<span class="k">def</span> <span class="nf">train_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">mnist</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">cifar</span> <span class="o">=</span> <span class="n">CIFAR</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">mnist_loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">mnist</span><span class="p">,</span> <span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="n">cifar_loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">cifar</span><span class="p">,</span> <span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">shuffle</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="c1"># each batch will be a dict of tensors: {&#39;mnist&#39;: batch_mnist, &#39;cifar&#39;: batch_cifar}</span>
<span class="k">return</span> <span class="p">{</span><span class="s1">&#39;mnist&#39;</span><span class="p">:</span> <span class="n">mnist_loader</span><span class="p">,</span> <span class="s1">&#39;cifar&#39;</span><span class="p">:</span> <span class="n">cifar_loader</span><span class="p">}</span>
</pre></div>
</div>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ManualDataModule.val_dataloader">
<span class="sig-name descname"><span class="pre">val_dataloader</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.ManualDataModule.val_dataloader" title="Permalink to this definition"></a></dt>
<dd><p>Implement one or multiple PyTorch DataLoaders for validation.</p>
<p>The dataloader you return will not be reloaded unless you set
<a href="#id11"><span class="problematic" id="id12">:paramref:`~pytorch_lightning.trainer.Trainer.reload_dataloaders_every_n_epochs`</span></a> to
a positive integer.</p>
<p>It’s recommended that all data downloads and preparation happen in <a class="reference internal" href="#reagent.data.ManualDataModule.prepare_data" title="reagent.data.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a>.</p>
<ul class="simple">
<li><p><code class="xref py py-meth docutils literal notranslate"><span class="pre">fit()</span></code></p></li>
<li><p>…</p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.prepare_data" title="reagent.data.ManualDataModule.prepare_data"><code class="xref py py-meth docutils literal notranslate"><span class="pre">prepare_data()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.train_dataloader" title="reagent.data.ManualDataModule.train_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">train_dataloader()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.val_dataloader" title="reagent.data.ManualDataModule.val_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">val_dataloader()</span></code></a></p></li>
<li><p><a class="reference internal" href="#reagent.data.ManualDataModule.test_dataloader" title="reagent.data.ManualDataModule.test_dataloader"><code class="xref py py-meth docutils literal notranslate"><span class="pre">test_dataloader()</span></code></a></p></li>
</ul>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Lightning adds the correct sampler for distributed and arbitrary hardware
There is no need to set it yourself.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Returns</dt>
<dd class="field-odd"><p>A <code class="xref py py-class docutils literal notranslate"><span class="pre">torch.utils.data.DataLoader</span></code> or a sequence of them specifying validation samples.</p>
</dd>
</dl>
<p>Examples:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">val_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="n">transform</span> <span class="o">=</span> <span class="n">transforms</span><span class="o">.</span><span class="n">Compose</span><span class="p">([</span><span class="n">transforms</span><span class="o">.</span><span class="n">ToTensor</span><span class="p">(),</span>
<span class="n">transforms</span><span class="o">.</span><span class="n">Normalize</span><span class="p">((</span><span class="mf">0.5</span><span class="p">,),</span> <span class="p">(</span><span class="mf">1.0</span><span class="p">,))])</span>
<span class="n">dataset</span> <span class="o">=</span> <span class="n">MNIST</span><span class="p">(</span><span class="n">root</span><span class="o">=</span><span class="s1">&#39;/path/to/mnist/&#39;</span><span class="p">,</span> <span class="n">train</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span>
<span class="n">transform</span><span class="o">=</span><span class="n">transform</span><span class="p">,</span> <span class="n">download</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="n">loader</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">utils</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">DataLoader</span><span class="p">(</span>
<span class="n">dataset</span><span class="o">=</span><span class="n">dataset</span><span class="p">,</span>
<span class="n">batch_size</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">batch_size</span><span class="p">,</span>
<span class="n">shuffle</span><span class="o">=</span><span class="kc">False</span>
<span class="p">)</span>
<span class="k">return</span> <span class="n">loader</span>
<span class="c1"># can also return multiple dataloaders</span>
<span class="k">def</span> <span class="nf">val_dataloader</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="k">return</span> <span class="p">[</span><span class="n">loader_a</span><span class="p">,</span> <span class="n">loader_b</span><span class="p">,</span> <span class="o">...</span><span class="p">,</span> <span class="n">loader_n</span><span class="p">]</span>
</pre></div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>If you don’t need a validation dataset and a <code class="xref py py-meth docutils literal notranslate"><span class="pre">validation_step()</span></code>, you don’t need to
implement this method.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>In the case where you return multiple validation dataloaders, the <code class="xref py py-meth docutils literal notranslate"><span class="pre">validation_step()</span></code>
will have an argument <code class="docutils literal notranslate"><span class="pre">dataloader_idx</span></code> which matches the order here.</p>
</div>
</dd></dl>
</dd></dl>
<dl class="py class">
<dt class="sig sig-object py" id="reagent.data.ReAgentDataModule">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">reagent.data.</span></span><span class="sig-name descname"><span class="pre">ReAgentDataModule</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="o"><span class="pre">*</span></span><span class="n"><span class="pre">args</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Any</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">kwargs</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Any</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#reagent.data.ReAgentDataModule" title="Permalink to this definition"></a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">pytorch_lightning.core.datamodule.LightningDataModule</span></code></p>
<dl class="py method">
<dt class="sig sig-object py" id="reagent.data.ReAgentDataModule.get_normalization_data_map">
<em class="property"><span class="pre">abstract</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">get_normalization_data_map</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">keys</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">List</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">]</span></span></span><span class="w"> </span><span class="o"><span class="pre">=</span></span><span class="w"> </span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">Dict</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="reagent.core.html#reagent.core.parameters.NormalizationData" title="reagent.core.parameters.NormalizationData"><span class="pre">reagent.core.parameters.NormalizationData</span></a><span class="p"><span class="pre">]</span></span></span></span><a class="headerlink" href="#reagent.data.ReAgentDataModule.get_normalization_data_map" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
</dd></dl>
</section>
</section>
</div>
</div>
<footer><div class="rst-footer-buttons" role="navigation" aria-label="Footer">
<a href="reagent.core.html" class="btn btn-neutral float-left" title="reagent.core package" accesskey="p" rel="prev"><span class="fa fa-arrow-circle-left" aria-hidden="true"></span> Previous</a>
<a href="reagent.gym.html" class="btn btn-neutral float-right" title="reagent.gym package" 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 2022, Meta Platforms, Inc.</p>
</div>
Built with <a href="https://www.sphinx-doc.org/">Sphinx</a> using a
<a href="https://github.com/readthedocs/sphinx_rtd_theme">theme</a>
provided by <a href="https://readthedocs.org">Read the Docs</a>.
</footer>
</div>
</div>
</section>
</div>
<script>
jQuery(function () {
SphinxRtdTheme.Navigation.enable(true);
});
</script>
</body>
</html>
Reference in New Issue View Git Blame Copy Permalink