Register
Login
Resources
Docs Blog Datasets Glossary Case Studies Tutorials & Webinars
Product
Data Engine LLMs Platform Enterprise
Pricing Explore
Connect to our Discord channel
Resources  
Docs Blog Datasets Glossary Case Studies Tutorials & Webinars
Explore Pricing
Product  
Data Engine LLMs Platform Enterprise
Connect to Dagshub Discord channel
Register
Login

Deci-AI
/
super-gradients
connected to https://github.com/Deci-AI/super-gradients.git

Watch 4
Issues 62 Pull Requests 44 Discussions 0

#572 new generated docs

Merged
Ghost merged 1 commits into Deci-AI:master from deci-ai:feature/SG-000_new_generated_docs
Conversation 0 Files changed 177 Commits 1
File Compare
super-gradients
/
docs
/
super_gradients.training.html
View left file View right file
Only showing up to 1000 lines per file, please use a local Git client to see the full diff.
Some lines were truncated since they exceed the maximum allowed length of 500, please use a local Git client to see the full diff.
@@ -1,13 +1,13 @@
 
 <!DOCTYPE html>
 
 <!DOCTYPE html>
 
 <html class="writer-html5" lang="en" >
 
 <html class="writer-html5" lang="en" >
 
 <head>
 
 <head>
 
-  <meta charset="utf-8" /><meta name="generator" content="Docutils 0.17.1: http://docutils.sourceforge.net/" />
 
-
 
+  <meta charset="utf-8" />
 
   <meta name="viewport" content="width=device-width, initial-scale=1.0" />
 
   <meta name="viewport" content="width=device-width, initial-scale=1.0" />
 
-  <title>Training package &mdash; SuperGradients 1.0 documentation</title>
 
+  <title>Training package &mdash; SuperGradients 3.0.3 documentation</title>
 
       <link rel="stylesheet" href="_static/pygments.css" type="text/css" />
 
       <link rel="stylesheet" href="_static/pygments.css" type="text/css" />
 
       <link rel="stylesheet" href="_static/css/theme.css" type="text/css" />
 
       <link rel="stylesheet" href="_static/css/theme.css" type="text/css" />
 
       <link rel="stylesheet" href="_static/graphviz.css" type="text/css" />
 
       <link rel="stylesheet" href="_static/graphviz.css" type="text/css" />
 
+      <link rel="stylesheet" href="_static/custom.css" type="text/css" />
 
   <!--[if lt IE 9]>
 
   <!--[if lt IE 9]>
 
     <script src="_static/js/html5shiv.min.js"></script>
 
     <script src="_static/js/html5shiv.min.js"></script>
 
   <![endif]-->
 
   <![endif]-->
 
@@ -15,11 +15,12 @@
 
         <script data-url_root="./" id="documentation_options" src="_static/documentation_options.js"></script>
 
         <script data-url_root="./" id="documentation_options" src="_static/documentation_options.js"></script>
 
         <script src="_static/jquery.js"></script>
 
         <script src="_static/jquery.js"></script>
 
         <script src="_static/underscore.js"></script>
 
         <script src="_static/underscore.js"></script>
 
+        <script src="_static/_sphinx_javascript_frameworks_compat.js"></script>
 
         <script src="_static/doctools.js"></script>
 
         <script src="_static/doctools.js"></script>
 
+        <script src="_static/sphinx_highlight.js"></script>
 
     <script src="_static/js/theme.js"></script>
 
     <script src="_static/js/theme.js"></script>
 
     <link rel="index" title="Index" href="genindex.html" />
 
     <link rel="index" title="Index" href="genindex.html" />
 
     <link rel="search" title="Search" href="search.html" />
 
     <link rel="search" title="Search" href="search.html" />
 
-    <link rel="next" title="What is SuperGradients?" href="user_guide.html" />
 
     <link rel="prev" title="Common package" href="super_gradients.common.html" /> 
     <link rel="prev" title="Common package" href="super_gradients.common.html" /> 
 </head>
 
 </head>
 
 
 
@@ -38,42 +39,568 @@
 
   </form>
 
   </form>
 
 </div>
 
 </div>
 
         </div><div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="Navigation menu">
 
         </div><div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="Navigation menu">
 
-              <p class="caption"><span class="caption-text">Welcome To SuperGradients</span></p>
 
+              <p class="caption" role="heading"><span class="caption-text">Welcome To SuperGradients</span></p>
 
 <ul>
 
 <ul>
 
-<li class="toctree-l1"><a class="reference internal" href="welcome.html">Fill our 4-question quick survey! We will raffle free SuperGradients swag between those who will participate -&gt; Fill Survey</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="welcome.html#supergradients">SuperGradients</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html">Version 3 is out! Notebooks have been updated!</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#build-with-supergradients">Build with SuperGradients</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#quick-installation">Quick Installation</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#what-s-new">What’s New</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#coming-soon">Coming soon</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#table-of-content">Table of Content</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#getting-started">Getting Started</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#advanced-features">Advanced Features</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#installation-methods">Installation Methods</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#implemented-model-architectures">Implemented Model Architectures</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#documentation">Documentation</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#contributing">Contributing</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#citation">Citation</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#community">Community</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#license">License</a></li>
 
+<li class="toctree-l1"><a class="reference internal" href="welcome.html#deci-platform">Deci Platform</a></li>
 
 </ul>
 
 </ul>
 
-<p class="caption"><span class="caption-text">Technical Documentation</span></p>
 
+<p class="caption" role="heading"><span class="caption-text">Technical Documentation</span></p>
 
 <ul class="current">
 
 <ul class="current">
 
 <li class="toctree-l1"><a class="reference internal" href="super_gradients.common.html">Common package</a></li>
 
 <li class="toctree-l1"><a class="reference internal" href="super_gradients.common.html">Common package</a></li>
 
 <li class="toctree-l1 current"><a class="current reference internal" href="#">Training package</a><ul>
 
 <li class="toctree-l1 current"><a class="current reference internal" href="#">Training package</a><ul>
 
-<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training">super_gradients.training module</a></li>
 
-<li class="toctree-l2"><a class="reference internal" href="#super-gradients-training-datasets-module">super_gradients.training.datasets module</a></li>
 
+<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training">super_gradients.training module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.DataAugmentation"><code class="docutils literal notranslate"><span class="pre">DataAugmentation</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.DataAugmentation.to_tensor"><code class="docutils literal notranslate"><span class="pre">DataAugmentation.to_tensor()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.DataAugmentation.normalize"><code class="docutils literal notranslate"><span class="pre">DataAugmentation.normalize()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.DataAugmentation.cutout"><code class="docutils literal notranslate"><span class="pre">DataAugmentation.cutout()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.Trainer"><code class="docutils literal notranslate"><span class="pre">Trainer</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.train"><code class="docutils literal notranslate"><span class="pre">Trainer.train()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.predict"><code class="docutils literal notranslate"><span class="pre">Trainer.predict()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.train_from_config"><code class="docutils literal notranslate"><span class="pre">Trainer.train_from_config()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.resume_experiment"><code class="docutils literal notranslate"><span class="pre">Trainer.resume_experiment()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.evaluate_from_recipe"><code class="docutils literal notranslate"><span class="pre">Trainer.evaluate_from_recipe()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.evaluate_checkpoint"><code class="docutils literal notranslate"><span class="pre">Trainer.evaluate_checkpoint()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#id0"><code class="docutils literal notranslate"><span class="pre">Trainer.train()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.get_arch_params"><code class="docutils literal notranslate"><span class="pre">Trainer.get_arch_params</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.get_structure"><code class="docutils literal notranslate"><span class="pre">Trainer.get_structure</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.get_architecture"><code class="docutils literal notranslate"><span class="pre">Trainer.get_architecture</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.set_experiment_name"><code class="docutils literal notranslate"><span class="pre">Trainer.set_experiment_name()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.get_module"><code class="docutils literal notranslate"><span class="pre">Trainer.get_module</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.set_module"><code class="docutils literal notranslate"><span class="pre">Trainer.set_module()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.test"><code class="docutils literal notranslate"><span class="pre">Trainer.test()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.evaluate"><code class="docutils literal notranslate"><span class="pre">Trainer.evaluate()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.get_net"><code class="docutils literal notranslate"><span class="pre">Trainer.get_net</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.set_net"><code class="docutils literal notranslate"><span class="pre">Trainer.set_net()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.set_ckpt_best_name"><code class="docutils literal notranslate"><span class="pre">Trainer.set_ckpt_best_name()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.Trainer.set_ema"><code class="docutils literal notranslate"><span class="pre">Trainer.set_ema()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.KDTrainer"><code class="docutils literal notranslate"><span class="pre">KDTrainer</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.KDTrainer.train_from_config"><code class="docutils literal notranslate"><span class="pre">KDTrainer.train_from_config()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.KDTrainer.train"><code class="docutils literal notranslate"><span class="pre">KDTrainer.train()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.MultiGPUMode"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.MultiGPUMode.OFF"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.OFF</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.MultiGPUMode.DATA_PARALLEL"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.DATA_PARALLEL</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.MultiGPUMode.DISTRIBUTED_DATA_PARALLEL"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.DISTRIBUTED_DATA_PARALLEL</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.MultiGPUMode.AUTO"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.AUTO</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.MultiGPUMode.dict"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.dict()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.StrictLoad"><code class="docutils literal notranslate"><span class="pre">StrictLoad</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.StrictLoad.OFF"><code class="docutils literal notranslate"><span class="pre">StrictLoad.OFF</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.StrictLoad.ON"><code class="docutils literal notranslate"><span class="pre">StrictLoad.ON</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.StrictLoad.NO_KEY_MATCHING"><code class="docutils literal notranslate"><span class="pre">StrictLoad.NO_KEY_MATCHING</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.EvaluationType"><code class="docutils literal notranslate"><span class="pre">EvaluationType</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.EvaluationType.TEST"><code class="docutils literal notranslate"><span class="pre">EvaluationType.TEST</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.EvaluationType.VALIDATION"><code class="docutils literal notranslate"><span class="pre">EvaluationType.VALIDATION</span></code></a></li>
 
+</ul>
 
+</li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l2"><a class="reference internal" href="#super-gradients-training-datasets-module">super_gradients.training.datasets module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.DataAugmentation"><code class="docutils literal notranslate"><span class="pre">DataAugmentation</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DataAugmentation.to_tensor"><code class="docutils literal notranslate"><span class="pre">DataAugmentation.to_tensor()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DataAugmentation.normalize"><code class="docutils literal notranslate"><span class="pre">DataAugmentation.normalize()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DataAugmentation.cutout"><code class="docutils literal notranslate"><span class="pre">DataAugmentation.cutout()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.ListDataset"><code class="docutils literal notranslate"><span class="pre">ListDataset</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.DirectoryDataSet"><code class="docutils literal notranslate"><span class="pre">DirectoryDataSet</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.SegmentationDataSet"><code class="docutils literal notranslate"><span class="pre">SegmentationDataSet</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.SegmentationDataSet.sample_loader"><code class="docutils literal notranslate"><span class="pre">SegmentationDataSet.sample_loader()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.SegmentationDataSet.sample_transform"><code class="docutils literal notranslate"><span class="pre">SegmentationDataSet.sample_transform()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.SegmentationDataSet.target_loader"><code class="docutils literal notranslate"><span class="pre">SegmentationDataSet.target_loader()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.SegmentationDataSet.target_transform"><code class="docutils literal notranslate"><span class="pre">SegmentationDataSet.target_transform()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.PascalVOC2012SegmentationDataSet"><code class="docutils literal notranslate"><span class="pre">PascalVOC2012SegmentationDataSet</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.PascalVOC2012SegmentationDataSet.IGNORE_LABEL"><code class="docutils literal notranslate"><span class="pre">PascalVOC2012SegmentationDataSet.IGNORE_LABEL</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.PascalVOC2012SegmentationDataSet.target_transform"><code class="docutils literal notranslate"><span class="pre">PascalVOC2012SegmentationDataSet.target_transform()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.PascalVOC2012SegmentationDataSet.decode_segmentation_mask"><code class="docutils literal notranslate"><span class="pre">PascalVOC2012SegmentationDataSet.decode_segmentation_mask()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.PascalAUG2012SegmentationDataSet"><code class="docutils literal notranslate"><span class="pre">PascalAUG2012SegmentationDataSet</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.PascalAUG2012SegmentationDataSet.target_loader"><code class="docutils literal notranslate"><span class="pre">PascalAUG2012SegmentationDataSet.target_loader()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.PascalVOCAndAUGUnifiedDataset"><code class="docutils literal notranslate"><span class="pre">PascalVOCAndAUGUnifiedDataset</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.PascalVOCAndAUGUnifiedDataset.datasets"><code class="docutils literal notranslate"><span class="pre">PascalVOCAndAUGUnifiedDataset.datasets</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.PascalVOCAndAUGUnifiedDataset.cumulative_sizes"><code class="docutils literal notranslate"><span class="pre">PascalVOCAndAUGUnifiedDataset.cumulative_sizes</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.CoCoSegmentationDataSet"><code class="docutils literal notranslate"><span class="pre">CoCoSegmentationDataSet</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.CoCoSegmentationDataSet.target_loader"><code class="docutils literal notranslate"><span class="pre">CoCoSegmentationDataSet.target_loader()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.DetectionDataset"><code class="docutils literal notranslate"><span class="pre">DetectionDataset</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DetectionDataset.get_random_item"><code class="docutils literal notranslate"><span class="pre">DetectionDataset.get_random_item()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DetectionDataset.get_sample"><code class="docutils literal notranslate"><span class="pre">DetectionDataset.get_sample()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DetectionDataset.get_resized_image"><code class="docutils literal notranslate"><span class="pre">DetectionDataset.get_resized_image()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DetectionDataset.apply_transforms"><code class="docutils literal notranslate"><span class="pre">DetectionDataset.apply_transforms()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DetectionDataset.get_random_samples"><code class="docutils literal notranslate"><span class="pre">DetectionDataset.get_random_samples()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DetectionDataset.get_random_sample"><code class="docutils literal notranslate"><span class="pre">DetectionDataset.get_random_sample()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DetectionDataset.output_target_format"><code class="docutils literal notranslate"><span class="pre">DetectionDataset.output_target_format</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.DetectionDataset.plot"><code class="docutils literal notranslate"><span class="pre">DetectionDataset.plot()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.COCODetectionDataset"><code class="docutils literal notranslate"><span class="pre">COCODetectionDataset</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.PascalVOCDetectionDataset"><code class="docutils literal notranslate"><span class="pre">PascalVOCDetectionDataset</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.PascalVOCDetectionDataset.download"><code class="docutils literal notranslate"><span class="pre">PascalVOCDetectionDataset.download()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.ImageNetDataset"><code class="docutils literal notranslate"><span class="pre">ImageNetDataset</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.Cifar10"><code class="docutils literal notranslate"><span class="pre">Cifar10</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.Cifar100"><code class="docutils literal notranslate"><span class="pre">Cifar100</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.datasets.SuperviselyPersonsDataset"><code class="docutils literal notranslate"><span class="pre">SuperviselyPersonsDataset</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.datasets.SuperviselyPersonsDataset.CLASS_LABELS"><code class="docutils literal notranslate"><span class="pre">SuperviselyPersonsDataset.CLASS_LABELS</span></code></a></li>
 
+</ul>
 
+</li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l2"><a class="reference internal" href="#super-gradients-training-dataloaders-module">super_gradients.training.dataloaders module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.coco2017_train"><code class="docutils literal notranslate"><span class="pre">coco2017_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.coco2017_val"><code class="docutils literal notranslate"><span class="pre">coco2017_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.coco2017_train_yolox"><code class="docutils literal notranslate"><span class="pre">coco2017_train_yolox()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.coco2017_val_yolox"><code class="docutils literal notranslate"><span class="pre">coco2017_val_yolox()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.coco2017_train_ssd_lite_mobilenet_v2"><code class="docutils literal notranslate"><span class="pre">coco2017_train_ssd_lite_mobilenet_v2()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.coco2017_val_ssd_lite_mobilenet_v2"><code class="docutils literal notranslate"><span class="pre">coco2017_val_ssd_lite_mobilenet_v2()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_train"><code class="docutils literal notranslate"><span class="pre">imagenet_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_val"><code class="docutils literal notranslate"><span class="pre">imagenet_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_efficientnet_train"><code class="docutils literal notranslate"><span class="pre">imagenet_efficientnet_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_efficientnet_val"><code class="docutils literal notranslate"><span class="pre">imagenet_efficientnet_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_mobilenetv2_train"><code class="docutils literal notranslate"><span class="pre">imagenet_mobilenetv2_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_mobilenetv2_val"><code class="docutils literal notranslate"><span class="pre">imagenet_mobilenetv2_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_mobilenetv3_train"><code class="docutils literal notranslate"><span class="pre">imagenet_mobilenetv3_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_mobilenetv3_val"><code class="docutils literal notranslate"><span class="pre">imagenet_mobilenetv3_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_regnetY_train"><code class="docutils literal notranslate"><span class="pre">imagenet_regnetY_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_regnetY_val"><code class="docutils literal notranslate"><span class="pre">imagenet_regnetY_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_resnet50_train"><code class="docutils literal notranslate"><span class="pre">imagenet_resnet50_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_resnet50_val"><code class="docutils literal notranslate"><span class="pre">imagenet_resnet50_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_resnet50_kd_train"><code class="docutils literal notranslate"><span class="pre">imagenet_resnet50_kd_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_resnet50_kd_val"><code class="docutils literal notranslate"><span class="pre">imagenet_resnet50_kd_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_vit_base_train"><code class="docutils literal notranslate"><span class="pre">imagenet_vit_base_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.imagenet_vit_base_val"><code class="docutils literal notranslate"><span class="pre">imagenet_vit_base_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.tiny_imagenet_train"><code class="docutils literal notranslate"><span class="pre">tiny_imagenet_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.tiny_imagenet_val"><code class="docutils literal notranslate"><span class="pre">tiny_imagenet_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cifar10_train"><code class="docutils literal notranslate"><span class="pre">cifar10_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cifar10_val"><code class="docutils literal notranslate"><span class="pre">cifar10_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cifar100_train"><code class="docutils literal notranslate"><span class="pre">cifar100_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cifar100_val"><code class="docutils literal notranslate"><span class="pre">cifar100_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_train"><code class="docutils literal notranslate"><span class="pre">cityscapes_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_val"><code class="docutils literal notranslate"><span class="pre">cityscapes_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_stdc_seg50_train"><code class="docutils literal notranslate"><span class="pre">cityscapes_stdc_seg50_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_stdc_seg50_val"><code class="docutils literal notranslate"><span class="pre">cityscapes_stdc_seg50_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_stdc_seg75_train"><code class="docutils literal notranslate"><span class="pre">cityscapes_stdc_seg75_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_stdc_seg75_val"><code class="docutils literal notranslate"><span class="pre">cityscapes_stdc_seg75_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_regseg48_train"><code class="docutils literal notranslate"><span class="pre">cityscapes_regseg48_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_regseg48_val"><code class="docutils literal notranslate"><span class="pre">cityscapes_regseg48_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_ddrnet_train"><code class="docutils literal notranslate"><span class="pre">cityscapes_ddrnet_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.cityscapes_ddrnet_val"><code class="docutils literal notranslate"><span class="pre">cityscapes_ddrnet_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.coco_segmentation_train"><code class="docutils literal notranslate"><span class="pre">coco_segmentation_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.coco_segmentation_val"><code class="docutils literal notranslate"><span class="pre">coco_segmentation_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.pascal_aug_segmentation_train"><code class="docutils literal notranslate"><span class="pre">pascal_aug_segmentation_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.pascal_aug_segmentation_val"><code class="docutils literal notranslate"><span class="pre">pascal_aug_segmentation_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.pascal_voc_segmentation_train"><code class="docutils literal notranslate"><span class="pre">pascal_voc_segmentation_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.pascal_voc_segmentation_val"><code class="docutils literal notranslate"><span class="pre">pascal_voc_segmentation_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.supervisely_persons_train"><code class="docutils literal notranslate"><span class="pre">supervisely_persons_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.supervisely_persons_val"><code class="docutils literal notranslate"><span class="pre">supervisely_persons_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.pascal_voc_detection_train"><code class="docutils literal notranslate"><span class="pre">pascal_voc_detection_train()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.pascal_voc_detection_val"><code class="docutils literal notranslate"><span class="pre">pascal_voc_detection_val()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.get_data_loader"><code class="docutils literal notranslate"><span class="pre">get_data_loader()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.dataloaders.get"><code class="docutils literal notranslate"><span class="pre">get()</span></code></a></li>
 
+</ul>
 
+</li>
 
 <li class="toctree-l2"><a class="reference internal" href="#super-gradients-training-exceptions-module">super_gradients.training.exceptions module</a></li>
 
 <li class="toctree-l2"><a class="reference internal" href="#super-gradients-training-exceptions-module">super_gradients.training.exceptions module</a></li>
 
+<li class="toctree-l2"><a class="reference internal" href="#super-gradients-training-kd-trainer-module">super_gradients.training.kd_trainer module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.kd_trainer.KDTrainer"><code class="docutils literal notranslate"><span class="pre">KDTrainer</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.kd_trainer.KDTrainer.train_from_config"><code class="docutils literal notranslate"><span class="pre">KDTrainer.train_from_config()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.kd_trainer.KDTrainer.train"><code class="docutils literal notranslate"><span class="pre">KDTrainer.train()</span></code></a></li>
 
+</ul>
 
+</li>
 
+</ul>
 
+</li>
 
 <li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.legacy">super_gradients.training.legacy module</a></li>
 
 <li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.legacy">super_gradients.training.legacy module</a></li>
 
-<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.losses">super_gradients.training.losses_models module</a></li>
 
-<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.metrics">super_gradients.training.metrics module</a></li>
 
+<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.losses">super_gradients.training.losses_models module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.Losses"><code class="docutils literal notranslate"><span class="pre">Losses</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.CROSS_ENTROPY"><code class="docutils literal notranslate"><span class="pre">Losses.CROSS_ENTROPY</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.MSE"><code class="docutils literal notranslate"><span class="pre">Losses.MSE</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.R_SQUARED_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.R_SQUARED_LOSS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.SHELFNET_OHEM_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.SHELFNET_OHEM_LOSS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.SHELFNET_SE_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.SHELFNET_SE_LOSS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.YOLOX_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.YOLOX_LOSS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.YOLOX_FAST_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.YOLOX_FAST_LOSS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.SSD_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.SSD_LOSS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.STDC_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.STDC_LOSS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.BCE_DICE_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.BCE_DICE_LOSS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.KD_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.KD_LOSS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.Losses.DICE_CE_EDGE_LOSS"><code class="docutils literal notranslate"><span class="pre">Losses.DICE_CE_EDGE_LOSS</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.FocalLoss"><code class="docutils literal notranslate"><span class="pre">FocalLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.FocalLoss.reduction"><code class="docutils literal notranslate"><span class="pre">FocalLoss.reduction</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.FocalLoss.forward"><code class="docutils literal notranslate"><span class="pre">FocalLoss.forward()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.LabelSmoothingCrossEntropyLoss"><code class="docutils literal notranslate"><span class="pre">LabelSmoothingCrossEntropyLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.LabelSmoothingCrossEntropyLoss.forward"><code class="docutils literal notranslate"><span class="pre">LabelSmoothingCrossEntropyLoss.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.LabelSmoothingCrossEntropyLoss.ignore_index"><code class="docutils literal notranslate"><span class="pre">LabelSmoothingCrossEntropyLoss.ignore_index</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.LabelSmoothingCrossEntropyLoss.label_smoothing"><code class="docutils literal notranslate"><span class="pre">LabelSmoothingCrossEntropyLoss.label_smoothing</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.ShelfNetOHEMLoss"><code class="docutils literal notranslate"><span class="pre">ShelfNetOHEMLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.ShelfNetOHEMLoss.forward"><code class="docutils literal notranslate"><span class="pre">ShelfNetOHEMLoss.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.ShelfNetOHEMLoss.component_names"><code class="docutils literal notranslate"><span class="pre">ShelfNetOHEMLoss.component_names</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.ShelfNetOHEMLoss.reduction"><code class="docutils literal notranslate"><span class="pre">ShelfNetOHEMLoss.reduction</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.ShelfNetSemanticEncodingLoss"><code class="docutils literal notranslate"><span class="pre">ShelfNetSemanticEncodingLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.ShelfNetSemanticEncodingLoss.forward"><code class="docutils literal notranslate"><span class="pre">ShelfNetSemanticEncodingLoss.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.ShelfNetSemanticEncodingLoss.component_names"><code class="docutils literal notranslate"><span class="pre">ShelfNetSemanticEncodingLoss.component_names</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.ShelfNetSemanticEncodingLoss.ignore_index"><code class="docutils literal notranslate"><span class="pre">ShelfNetSemanticEncodingLoss.ignore_index</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.ShelfNetSemanticEncodingLoss.label_smoothing"><code class="docutils literal notranslate"><span class="pre">ShelfNetSemanticEncodingLoss.label_smoothing</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.strides"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.strides</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.num_classes"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.num_classes</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.use_l1"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.use_l1</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.center_sampling_radius"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.center_sampling_radius</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.iou_type"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.iou_type</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.component_names"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.component_names</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.forward"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.prepare_predictions"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.prepare_predictions()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.get_l1_target"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.get_l1_target()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.get_assignments"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.get_assignments()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.get_in_boxes_info"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.get_in_boxes_info()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.dynamic_k_matching"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.dynamic_k_matching()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXDetectionLoss.reduction"><code class="docutils literal notranslate"><span class="pre">YoloXDetectionLoss.reduction</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.YoloXFastDetectionLoss"><code class="docutils literal notranslate"><span class="pre">YoloXFastDetectionLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXFastDetectionLoss.reduction"><code class="docutils literal notranslate"><span class="pre">YoloXFastDetectionLoss.reduction</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.YoloXFastDetectionLoss.training"><code class="docutils literal notranslate"><span class="pre">YoloXFastDetectionLoss.training</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.RSquaredLoss"><code class="docutils literal notranslate"><span class="pre">RSquaredLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.RSquaredLoss.forward"><code class="docutils literal notranslate"><span class="pre">RSquaredLoss.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.RSquaredLoss.reduction"><code class="docutils literal notranslate"><span class="pre">RSquaredLoss.reduction</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.SSDLoss"><code class="docutils literal notranslate"><span class="pre">SSDLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.SSDLoss.component_names"><code class="docutils literal notranslate"><span class="pre">SSDLoss.component_names</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.SSDLoss.match_dboxes"><code class="docutils literal notranslate"><span class="pre">SSDLoss.match_dboxes()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.SSDLoss.forward"><code class="docutils literal notranslate"><span class="pre">SSDLoss.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.SSDLoss.reduction"><code class="docutils literal notranslate"><span class="pre">SSDLoss.reduction</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.BCEDiceLoss"><code class="docutils literal notranslate"><span class="pre">BCEDiceLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.BCEDiceLoss.loss_weights"><code class="docutils literal notranslate"><span class="pre">BCEDiceLoss.loss_weights</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.BCEDiceLoss.forward"><code class="docutils literal notranslate"><span class="pre">BCEDiceLoss.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.BCEDiceLoss.training"><code class="docutils literal notranslate"><span class="pre">BCEDiceLoss.training</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.KDLogitsLoss"><code class="docutils literal notranslate"><span class="pre">KDLogitsLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.KDLogitsLoss.component_names"><code class="docutils literal notranslate"><span class="pre">KDLogitsLoss.component_names</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.KDLogitsLoss.forward"><code class="docutils literal notranslate"><span class="pre">KDLogitsLoss.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.KDLogitsLoss.reduction"><code class="docutils literal notranslate"><span class="pre">KDLogitsLoss.reduction</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.losses.DiceCEEdgeLoss"><code class="docutils literal notranslate"><span class="pre">DiceCEEdgeLoss</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.DiceCEEdgeLoss.component_names"><code class="docutils literal notranslate"><span class="pre">DiceCEEdgeLoss.component_names</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.DiceCEEdgeLoss.forward"><code class="docutils literal notranslate"><span class="pre">DiceCEEdgeLoss.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.losses.DiceCEEdgeLoss.reduction"><code class="docutils literal notranslate"><span class="pre">DiceCEEdgeLoss.reduction</span></code></a></li>
 
+</ul>
 
+</li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.metrics">super_gradients.training.metrics module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.Metrics"><code class="docutils literal notranslate"><span class="pre">Metrics</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.ACCURACY"><code class="docutils literal notranslate"><span class="pre">Metrics.ACCURACY</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.TOP5"><code class="docutils literal notranslate"><span class="pre">Metrics.TOP5</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.DETECTION_METRICS"><code class="docutils literal notranslate"><span class="pre">Metrics.DETECTION_METRICS</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.DETECTION_METRICS_050_095"><code class="docutils literal notranslate"><span class="pre">Metrics.DETECTION_METRICS_050_095</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.DETECTION_METRICS_050"><code class="docutils literal notranslate"><span class="pre">Metrics.DETECTION_METRICS_050</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.DETECTION_METRICS_075"><code class="docutils literal notranslate"><span class="pre">Metrics.DETECTION_METRICS_075</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.IOU"><code class="docutils literal notranslate"><span class="pre">Metrics.IOU</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.BINARY_IOU"><code class="docutils literal notranslate"><span class="pre">Metrics.BINARY_IOU</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.DICE"><code class="docutils literal notranslate"><span class="pre">Metrics.DICE</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.BINARY_DICE"><code class="docutils literal notranslate"><span class="pre">Metrics.BINARY_DICE</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Metrics.PIXEL_ACCURACY"><code class="docutils literal notranslate"><span class="pre">Metrics.PIXEL_ACCURACY</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.accuracy"><code class="docutils literal notranslate"><span class="pre">accuracy()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.Accuracy"><code class="docutils literal notranslate"><span class="pre">Accuracy</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Accuracy.update"><code class="docutils literal notranslate"><span class="pre">Accuracy.update()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Accuracy.correct"><code class="docutils literal notranslate"><span class="pre">Accuracy.correct</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Accuracy.total"><code class="docutils literal notranslate"><span class="pre">Accuracy.total</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.Top5"><code class="docutils literal notranslate"><span class="pre">Top5</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Top5.update"><code class="docutils literal notranslate"><span class="pre">Top5.update()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Top5.compute"><code class="docutils literal notranslate"><span class="pre">Top5.compute()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.ToyTestClassificationMetric"><code class="docutils literal notranslate"><span class="pre">ToyTestClassificationMetric</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.ToyTestClassificationMetric.update"><code class="docutils literal notranslate"><span class="pre">ToyTestClassificationMetric.update()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.ToyTestClassificationMetric.compute"><code class="docutils literal notranslate"><span class="pre">ToyTestClassificationMetric.compute()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.num_cls"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.num_cls</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.post_prediction_callback"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.post_prediction_callback</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.normalize_targets"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.normalize_targets</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.iou_thresholds"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.iou_thresholds</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.recall_thresholds"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.recall_thresholds</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.score_threshold"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.score_threshold</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.top_k_predictions"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.top_k_predictions</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.dist_sync_on_step"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.dist_sync_on_step</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.update"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.update()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics.compute"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics.compute()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.PreprocessSegmentationMetricsArgs"><code class="docutils literal notranslate"><span class="pre">PreprocessSegmentationMetricsArgs</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.PixelAccuracy"><code class="docutils literal notranslate"><span class="pre">PixelAccuracy</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.PixelAccuracy.update"><code class="docutils literal notranslate"><span class="pre">PixelAccuracy.update()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.PixelAccuracy.compute"><code class="docutils literal notranslate"><span class="pre">PixelAccuracy.compute()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.IoU"><code class="docutils literal notranslate"><span class="pre">IoU</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.IoU.update"><code class="docutils literal notranslate"><span class="pre">IoU.update()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.IoU.confmat"><code class="docutils literal notranslate"><span class="pre">IoU.confmat</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.Dice"><code class="docutils literal notranslate"><span class="pre">Dice</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Dice.update"><code class="docutils literal notranslate"><span class="pre">Dice.update()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Dice.compute"><code class="docutils literal notranslate"><span class="pre">Dice.compute()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.Dice.confmat"><code class="docutils literal notranslate"><span class="pre">Dice.confmat</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.BinaryIOU"><code class="docutils literal notranslate"><span class="pre">BinaryIOU</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.BinaryIOU.compute"><code class="docutils literal notranslate"><span class="pre">BinaryIOU.compute()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.BinaryIOU.confmat"><code class="docutils literal notranslate"><span class="pre">BinaryIOU.confmat</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.BinaryIOU.training"><code class="docutils literal notranslate"><span class="pre">BinaryIOU.training</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.BinaryDice"><code class="docutils literal notranslate"><span class="pre">BinaryDice</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.BinaryDice.compute"><code class="docutils literal notranslate"><span class="pre">BinaryDice.compute()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.BinaryDice.confmat"><code class="docutils literal notranslate"><span class="pre">BinaryDice.confmat</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.metrics.BinaryDice.training"><code class="docutils literal notranslate"><span class="pre">BinaryDice.training</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics_050"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics_050</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics_075"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics_075</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.metrics.DetectionMetrics_050_095"><code class="docutils literal notranslate"><span class="pre">DetectionMetrics_050_095</span></code></a></li>
 
+</ul>
 
+</li>
 
 <li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.models">super_gradients.training.models module</a></li>
 
 <li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.models">super_gradients.training.models module</a></li>
 
-<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.sg_model">super_gradients.training.sg_model module</a></li>
 
-<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.utils">super_gradients.training.utils module</a></li>
 
-<li class="toctree-l2"><a class="reference internal" href="#module-contents">Module contents</a></li>
 
+<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.sg_trainer">super_gradients.training.sg_model module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer"><code class="docutils literal notranslate"><span class="pre">Trainer</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.train"><code class="docutils literal notranslate"><span class="pre">Trainer.train()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.predict"><code class="docutils literal notranslate"><span class="pre">Trainer.predict()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.train_from_config"><code class="docutils literal notranslate"><span class="pre">Trainer.train_from_config()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.resume_experiment"><code class="docutils literal notranslate"><span class="pre">Trainer.resume_experiment()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.evaluate_from_recipe"><code class="docutils literal notranslate"><span class="pre">Trainer.evaluate_from_recipe()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.evaluate_checkpoint"><code class="docutils literal notranslate"><span class="pre">Trainer.evaluate_checkpoint()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#id3"><code class="docutils literal notranslate"><span class="pre">Trainer.train()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.get_arch_params"><code class="docutils literal notranslate"><span class="pre">Trainer.get_arch_params</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.get_structure"><code class="docutils literal notranslate"><span class="pre">Trainer.get_structure</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.get_architecture"><code class="docutils literal notranslate"><span class="pre">Trainer.get_architecture</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.set_experiment_name"><code class="docutils literal notranslate"><span class="pre">Trainer.set_experiment_name()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.get_module"><code class="docutils literal notranslate"><span class="pre">Trainer.get_module</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.set_module"><code class="docutils literal notranslate"><span class="pre">Trainer.set_module()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.test"><code class="docutils literal notranslate"><span class="pre">Trainer.test()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.evaluate"><code class="docutils literal notranslate"><span class="pre">Trainer.evaluate()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.get_net"><code class="docutils literal notranslate"><span class="pre">Trainer.get_net</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.set_net"><code class="docutils literal notranslate"><span class="pre">Trainer.set_net()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.set_ckpt_best_name"><code class="docutils literal notranslate"><span class="pre">Trainer.set_ckpt_best_name()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.Trainer.set_ema"><code class="docutils literal notranslate"><span class="pre">Trainer.set_ema()</span></code></a></li>
 
 </ul>
 
 </ul>
 
 </li>
 
 </li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.sg_trainer.MultiGPUMode"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.MultiGPUMode.OFF"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.OFF</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.MultiGPUMode.DATA_PARALLEL"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.DATA_PARALLEL</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.MultiGPUMode.DISTRIBUTED_DATA_PARALLEL"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.DISTRIBUTED_DATA_PARALLEL</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.MultiGPUMode.AUTO"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.AUTO</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.MultiGPUMode.dict"><code class="docutils literal notranslate"><span class="pre">MultiGPUMode.dict()</span></code></a></li>
 
 </ul>
 
 </ul>
 
-<p class="caption"><span class="caption-text">User Guide</span></p>
 
-<ul>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html">What is SuperGradients?</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#introducing-the-supergradients-library">Introducing the SuperGradients library</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#installation">Installation</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#integrating-your-training-code-complete-walkthrough">Integrating Your Training Code - Complete Walkthrough</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#training-parameters">Training Parameters</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#logs-and-checkpoints">Logs and Checkpoints</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#dataset-parameters">Dataset Parameters</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#network-architectures">Network Architectures</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#pretrained-models">Pretrained Models</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#how-to-reproduce-our-training-recipes">How To Reproduce Our Training Recipes</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#professional-tools-integration">Professional Tools Integration</a></li>
 
-<li class="toctree-l1"><a class="reference internal" href="user_guide.html#supergradients-faq">SuperGradients FAQ</a></li>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.sg_trainer.StrictLoad"><code class="docutils literal notranslate"><span class="pre">StrictLoad</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.StrictLoad.OFF"><code class="docutils literal notranslate"><span class="pre">StrictLoad.OFF</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.StrictLoad.ON"><code class="docutils literal notranslate"><span class="pre">StrictLoad.ON</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.sg_trainer.StrictLoad.NO_KEY_MATCHING"><code class="docutils literal notranslate"><span class="pre">StrictLoad.NO_KEY_MATCHING</span></code></a></li>
 
+</ul>
 
+</li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l2"><a class="reference internal" href="#super-gradients-training-training-hyperparams-module">super_gradients.training.training_hyperparams module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.cifar10_resnet_train_params"><code class="docutils literal notranslate"><span class="pre">cifar10_resnet_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.cityscapes_ddrnet_train_params"><code class="docutils literal notranslate"><span class="pre">cityscapes_ddrnet_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.cityscapes_regseg48_train_params"><code class="docutils literal notranslate"><span class="pre">cityscapes_regseg48_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.cityscapes_stdc_base_train_params"><code class="docutils literal notranslate"><span class="pre">cityscapes_stdc_base_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.cityscapes_stdc_seg50_train_params"><code class="docutils literal notranslate"><span class="pre">cityscapes_stdc_seg50_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.cityscapes_stdc_seg75_train_params"><code class="docutils literal notranslate"><span class="pre">cityscapes_stdc_seg75_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.coco2017_ssd_lite_mobilenet_v2_train_params"><code class="docutils literal notranslate"><span class="pre">coco2017_ssd_lite_mobilenet_v2_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.coco2017_yolox_train_params"><code class="docutils literal notranslate"><span class="pre">coco2017_yolox_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.coco_segmentation_shelfnet_lw_train_params"><code class="docutils literal notranslate"><span class="pre">coco_segmentation_shelfnet_lw_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_efficientnet_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_efficientnet_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_mobilenetv2_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_mobilenetv2_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_mobilenetv3_base_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_mobilenetv3_base_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_mobilenetv3_large_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_mobilenetv3_large_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_mobilenetv3_small_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_mobilenetv3_small_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_regnetY_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_regnetY_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_repvgg_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_repvgg_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_resnet50_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_resnet50_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_resnet50_kd_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_resnet50_kd_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_vit_base_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_vit_base_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.imagenet_vit_large_train_params"><code class="docutils literal notranslate"><span class="pre">imagenet_vit_large_train_params()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.training_hyperparams.get"><code class="docutils literal notranslate"><span class="pre">get()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l2"><a class="reference internal" href="#super-gradients-training-transforms-module">super_gradients.training.transforms module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.transforms.Transforms"><code class="docutils literal notranslate"><span class="pre">Transforms</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.SegRandomFlip"><code class="docutils literal notranslate"><span class="pre">Transforms.SegRandomFlip</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.SegResize"><code class="docutils literal notranslate"><span class="pre">Transforms.SegResize</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.SegRescale"><code class="docutils literal notranslate"><span class="pre">Transforms.SegRescale</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.SegRandomRescale"><code class="docutils literal notranslate"><span class="pre">Transforms.SegRandomRescale</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.SegRandomRotate"><code class="docutils literal notranslate"><span class="pre">Transforms.SegRandomRotate</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.SegCropImageAndMask"><code class="docutils literal notranslate"><span class="pre">Transforms.SegCropImageAndMask</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.SegRandomGaussianBlur"><code class="docutils literal notranslate"><span class="pre">Transforms.SegRandomGaussianBlur</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.SegPadShortToCropSize"><code class="docutils literal notranslate"><span class="pre">Transforms.SegPadShortToCropSize</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.SegColorJitter"><code class="docutils literal notranslate"><span class="pre">Transforms.SegColorJitter</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.DetectionMosaic"><code class="docutils literal notranslate"><span class="pre">Transforms.DetectionMosaic</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.DetectionRandomAffine"><code class="docutils literal notranslate"><span class="pre">Transforms.DetectionRandomAffine</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.DetectionMixup"><code class="docutils literal notranslate"><span class="pre">Transforms.DetectionMixup</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.DetectionHSV"><code class="docutils literal notranslate"><span class="pre">Transforms.DetectionHSV</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.DetectionHorizontalFlip"><code class="docutils literal notranslate"><span class="pre">Transforms.DetectionHorizontalFlip</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.DetectionPaddedRescale"><code class="docutils literal notranslate"><span class="pre">Transforms.DetectionPaddedRescale</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.DetectionTargetsFormat"><code class="docutils literal notranslate"><span class="pre">Transforms.DetectionTargetsFormat</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.DetectionTargetsFormatTransform"><code class="docutils literal notranslate"><span class="pre">Transforms.DetectionTargetsFormatTransform</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomResizedCropAndInterpolation"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomResizedCropAndInterpolation</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandAugmentTransform"><code class="docutils literal notranslate"><span class="pre">Transforms.RandAugmentTransform</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.Lighting"><code class="docutils literal notranslate"><span class="pre">Transforms.Lighting</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomErase"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomErase</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.Compose"><code class="docutils literal notranslate"><span class="pre">Transforms.Compose</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.ToTensor"><code class="docutils literal notranslate"><span class="pre">Transforms.ToTensor</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.PILToTensor"><code class="docutils literal notranslate"><span class="pre">Transforms.PILToTensor</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.ConvertImageDtype"><code class="docutils literal notranslate"><span class="pre">Transforms.ConvertImageDtype</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.ToPILImage"><code class="docutils literal notranslate"><span class="pre">Transforms.ToPILImage</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.Normalize"><code class="docutils literal notranslate"><span class="pre">Transforms.Normalize</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.Resize"><code class="docutils literal notranslate"><span class="pre">Transforms.Resize</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.CenterCrop"><code class="docutils literal notranslate"><span class="pre">Transforms.CenterCrop</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.Pad"><code class="docutils literal notranslate"><span class="pre">Transforms.Pad</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.Lambda"><code class="docutils literal notranslate"><span class="pre">Transforms.Lambda</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomApply"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomApply</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomChoice"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomChoice</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomOrder"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomOrder</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomCrop"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomCrop</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomHorizontalFlip"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomHorizontalFlip</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomVerticalFlip"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomVerticalFlip</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomResizedCrop"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomResizedCrop</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.FiveCrop"><code class="docutils literal notranslate"><span class="pre">Transforms.FiveCrop</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.TenCrop"><code class="docutils literal notranslate"><span class="pre">Transforms.TenCrop</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.LinearTransformation"><code class="docutils literal notranslate"><span class="pre">Transforms.LinearTransformation</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.ColorJitter"><code class="docutils literal notranslate"><span class="pre">Transforms.ColorJitter</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomRotation"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomRotation</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomAffine"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomAffine</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.Grayscale"><code class="docutils literal notranslate"><span class="pre">Transforms.Grayscale</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomGrayscale"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomGrayscale</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomPerspective"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomPerspective</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomErasing"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomErasing</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.GaussianBlur"><code class="docutils literal notranslate"><span class="pre">Transforms.GaussianBlur</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.InterpolationMode"><code class="docutils literal notranslate"><span class="pre">Transforms.InterpolationMode</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomInvert"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomInvert</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomPosterize"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomPosterize</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomSolarize"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomSolarize</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomAdjustSharpness"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomAdjustSharpness</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomAutocontrast"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomAutocontrast</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.Transforms.RandomEqualize"><code class="docutils literal notranslate"><span class="pre">Transforms.RandomEqualize</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.transforms.DetectionMosaic"><code class="docutils literal notranslate"><span class="pre">DetectionMosaic</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionMosaic.input_dim"><code class="docutils literal notranslate"><span class="pre">DetectionMosaic.input_dim</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionMosaic.prob"><code class="docutils literal notranslate"><span class="pre">DetectionMosaic.prob</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionMosaic.enable_mosaic"><code class="docutils literal notranslate"><span class="pre">DetectionMosaic.enable_mosaic</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionMosaic.close"><code class="docutils literal notranslate"><span class="pre">DetectionMosaic.close()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.target_size"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.target_size</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.degrees"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.degrees</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.translate"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.translate</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.scales"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.scales</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.shear"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.shear</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.enable"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.enable</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.filter_box_candidates"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.filter_box_candidates</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.wh_thr"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.wh_thr</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.ar_thr"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.ar_thr</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.area_thr"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.area_thr</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionRandomAffine.close"><code class="docutils literal notranslate"><span class="pre">DetectionRandomAffine.close()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.transforms.DetectionHSV"><code class="docutils literal notranslate"><span class="pre">DetectionHSV</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.transforms.DetectionPaddedRescale"><code class="docutils literal notranslate"><span class="pre">DetectionPaddedRescale</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionPaddedRescale.input_dim"><code class="docutils literal notranslate"><span class="pre">DetectionPaddedRescale.input_dim</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionPaddedRescale.swap"><code class="docutils literal notranslate"><span class="pre">DetectionPaddedRescale.swap</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.transforms.DetectionTargetsFormatTransform"><code class="docutils literal notranslate"><span class="pre">DetectionTargetsFormatTransform</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionTargetsFormatTransform.output_format"><code class="docutils literal notranslate"><span class="pre">DetectionTargetsFormatTransform.output_format</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionTargetsFormatTransform.min_bbox_edge_size"><code class="docutils literal notranslate"><span class="pre">DetectionTargetsFormatTransform.min_bbox_edge_size</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.transforms.DetectionTargetsFormatTransform.max_targets"><code class="docutils literal notranslate"><span class="pre">DetectionTargetsFormatTransform.max_targets</span></code></a></li>
 
+</ul>
 
+</li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l2"><a class="reference internal" href="#module-super_gradients.training.utils">super_gradients.training.utils module</a><ul>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.Timer"><code class="docutils literal notranslate"><span class="pre">Timer</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.utils.Timer.start"><code class="docutils literal notranslate"><span class="pre">Timer.start()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.utils.Timer.stop"><code class="docutils literal notranslate"><span class="pre">Timer.stop()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.HpmStruct"><code class="docutils literal notranslate"><span class="pre">HpmStruct</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.utils.HpmStruct.set_schema"><code class="docutils literal notranslate"><span class="pre">HpmStruct.set_schema()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.utils.HpmStruct.override"><code class="docutils literal notranslate"><span class="pre">HpmStruct.override()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.utils.HpmStruct.to_dict"><code class="docutils literal notranslate"><span class="pre">HpmStruct.to_dict()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.utils.HpmStruct.validate"><code class="docutils literal notranslate"><span class="pre">HpmStruct.validate()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.WrappedModel"><code class="docutils literal notranslate"><span class="pre">WrappedModel</span></code></a><ul>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.utils.WrappedModel.forward"><code class="docutils literal notranslate"><span class="pre">WrappedModel.forward()</span></code></a></li>
 
+<li class="toctree-l4"><a class="reference internal" href="#super_gradients.training.utils.WrappedModel.training"><code class="docutils literal notranslate"><span class="pre">WrappedModel.training</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.convert_to_tensor"><code class="docutils literal notranslate"><span class="pre">convert_to_tensor()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.get_param"><code class="docutils literal notranslate"><span class="pre">get_param()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.tensor_container_to_device"><code class="docutils literal notranslate"><span class="pre">tensor_container_to_device()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.adapt_state_dict_to_fit_model_layer_names"><code class="docutils literal notranslate"><span class="pre">adapt_state_dict_to_fit_model_layer_names()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.raise_informative_runtime_error"><code class="docutils literal notranslate"><span class="pre">raise_informative_runtime_error()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.random_seed"><code class="docutils literal notranslate"><span class="pre">random_seed()</span></code></a></li>
 
+<li class="toctree-l3"><a class="reference internal" href="#super_gradients.training.utils.torch_version_is_greater_or_equal"><code class="docutils literal notranslate"><span class="pre">torch_version_is_greater_or_equal()</span></code></a></li>
 
+</ul>
 
+</li>
 
+<li class="toctree-l2"><a class="reference internal" href="#module-contents">Module contents</a></li>
 
+</ul>
 
+</li>
 
 </ul>
 
 </ul>
 
 
 
         </div>
 
         </div>
 
@@ -100,9 +627,9 @@
 
           <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
 
           <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
 
            <div itemprop="articleBody">
 
            <div itemprop="articleBody">
 
              
              
-  <section id="training-package">
 
-<h1>Training package<a class="headerlink" href="#training-package" title="Permalink to this headline"></a></h1>
 
-<table class="longtable docutils align-default">
 
+  <div class="section" id="training-package">
 
+<h1>Training package<a class="headerlink" href="#training-package" title="Permalink to this heading"></a></h1>
 
+<table class="autosummary longtable docutils align-default">
 
 <colgroup>
 
 <colgroup>
 
 <col style="width: 10%" />
 
 <col style="width: 10%" />
 
 <col style="width: 90%" />
 
 <col style="width: 90%" />
 
@@ -110,61 +637,43 @@
 
 <tbody>
 
 <tbody>
 
 </tbody>
 
 </tbody>
 
 </table>
 
 </table>
 
-<section id="module-super_gradients.training">
 
-<span id="super-gradients-training-module"></span><h2>super_gradients.training module<a class="headerlink" href="#module-super_gradients.training" title="Permalink to this headline"></a></h2>
 
+<div class="section" id="module-super_gradients.training">
 
+<span id="super-gradients-training-module"></span><h2>super_gradients.training module<a class="headerlink" href="#module-super_gradients.training" title="Permalink to this heading"></a></h2>
 
 <dl class="py class">
 
 <dl class="py class">
 
 <dt class="sig sig-object py" id="super_gradients.training.DataAugmentation">
 
 <dt class="sig sig-object py" id="super_gradients.training.DataAugmentation">
 
-<em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">super_gradients.training.</span></span><span class="sig-name descname"><span class="pre">DataAugmentation</span></span><a class="reference internal" href="_modules/super_gradients/training/datasets/data_augmentation.html#DataAugmentation"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.DataAugmentation" t
 
+<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">super_gradients.training.</span></span><span class="sig-name descname"><span class="pre">DataAugmentation</span></span><a class="reference internal" href="_modules/super_gradients/training/datasets/data_augmentation.html#DataAugmentation"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.train
 
 <dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">object</span></code></p>
 
 <dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">object</span></code></p>
 
 <dl class="py method">
 
 <dl class="py method">
 
 <dt class="sig sig-object py" id="super_gradients.training.DataAugmentation.to_tensor">
 
 <dt class="sig sig-object py" id="super_gradients.training.DataAugmentation.to_tensor">
 
-<em class="property"><span class="pre">static</span> </em><span class="sig-name descname"><span class="pre">to_tensor</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/datasets/data_augmentation.html#DataAugmentation.to_tensor"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.DataAugmentation.to_tensor" title="Permalink to 
+<em class="property"><span class="pre">static</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">to_tensor</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/datasets/data_augmentation.html#DataAugmentation.to_tensor"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.DataAugmentation.to_tenso
 
 <dd></dd></dl>
 
 <dd></dd></dl>
 
 
 
 <dl class="py method">
 
 <dl class="py method">
 
 <dt class="sig sig-object py" id="super_gradients.training.DataAugmentation.normalize">
 
 <dt class="sig sig-object py" id="super_gradients.training.DataAugmentation.normalize">
 
-<em class="property"><span class="pre">static</span> </em><span class="sig-name descname"><span class="pre">normalize</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">mean</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">std</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/datasets/data_augmentation.html#DataAugmentation.normalize"><span class="vi
 
+<em class="property"><span class="pre">static</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">normalize</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">mean</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">std</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/datasets/data_augmentation.html#DataAugmentation.nor
 
 <dd></dd></dl>
 
 <dd></dd></dl>
 
 
 
 <dl class="py method">
 
 <dl class="py method">
 
 <dt class="sig sig-object py" id="super_gradients.training.DataAugmentation.cutout">
 
 <dt class="sig sig-object py" id="super_gradients.training.DataAugmentation.cutout">
 
-<em class="property"><span class="pre">static</span> </em><span class="sig-name descname"><span class="pre">cutout</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">mask_size</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">p</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">1</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">cutout_in
 
+<em class="property"><span class="pre">static</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">cutout</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">mask_size</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">p</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">1</span></span></em>, <em class="sig-param"><span class="n"><spa
 
 <dd></dd></dl>
 
 <dd></dd></dl>
 
 
 
 </dd></dl>
 
 </dd></dl>
 
 
 
 <dl class="py class">
 
 <dl class="py class">
 
-<dt class="sig sig-object py" id="super_gradients.training.TestDatasetInterface">
 
-<em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">super_gradients.training.</span></span><span class="sig-name descname"><span class="pre">TestDatasetInterface</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">trainset</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">dataset_params</span></span><span class="o"><span class="pre">=</span></span><span class
 
-<dd><p>Bases: <a class="reference internal" href="super_gradients.training.datasets.dataset_interfaces.html#super_gradients.training.datasets.dataset_interfaces.dataset_interface.DatasetInterface" title="super_gradients.training.datasets.dataset_interfaces.dataset_interface.DatasetInterface"><code class="xref py py-class docutils literal notranslate"><span class="pre">super_gradients.training.datasets.dataset_interfaces.dataset_interface.DatasetInterface</span></code></a></p>
 
-<dl class="py method">
 
-<dt class="sig sig-object py" id="super_gradients.training.TestDatasetInterface.get_data_loaders">
 
-<span class="sig-name descname"><span class="pre">get_data_loaders</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">batch_size_factor</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">1</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">num_workers</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">8</span
 
-<dd><p>Get self.train_loader, self.val_loader, self.test_loader, self.classes.</p>
 
-<p>If the data loaders haven’t been initialized yet, build them first.</p>
 
-<dl class="field-list simple">
 
-<dt class="field-odd">Parameters</dt>
 
-<dd class="field-odd"><p><strong>kwargs</strong> – kwargs are passed to build_data_loaders.</p>
 
-</dd>
 
-</dl>
 
-</dd></dl>
 
-
 
-</dd></dl>
 
-
 
-<dl class="py class">
 
-<dt class="sig sig-object py" id="super_gradients.training.SgModel">
 
-<em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">super_gradients.training.</span></span><span class="sig-name descname"><span class="pre">SgModel</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="pre">experiment_name:</span> <span class="pre">str</span></em>, <em class="sig-param"><span class="pre">device:</span> <span class="pre">Optional[str]</span> <span class="pre">=</span> <span class="pre">None
 
+<dt class="sig sig-object py" id="super_gradients.training.Trainer">
 
+<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">super_gradients.training.</span></span><span class="sig-name descname"><span class="pre">Trainer</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">experiment_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
 
 <dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">object</span></code></p>
 
 <dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">object</span></code></p>
 
 <p>SuperGradient Model - Base Class for Sg Models</p>
 
 <p>SuperGradient Model - Base Class for Sg Models</p>
 
 <dl class="py method">
 
 <dl class="py method">
 
-<dt class="sig sig-object py" id="super_gradients.training.SgModel.train">
 
-<span class="sig-name descname"><span class="pre">train</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">max_epochs</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">int</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">initial_epoch</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">int</span></span></em>, <em class="sig-para
 
+<dt class="sig sig-object py" id="super_gradients.training.Trainer.train">
 
+<span class="sig-name descname"><span class="pre">train</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">max_epochs</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">initial_epoch</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pr
 
 <dd><p>the main function used for the training, h.p. updating, logging etc.</p>
 
 <dd><p>the main function used for the training, h.p. updating, logging etc.</p>
 
 </dd></dl>
 
 </dd></dl>
 
 
 
 <dl class="py method">
 
 <dl class="py method">
 
-<dt class="sig sig-object py" id="super_gradients.training.SgModel.predict">
 
-<span class="sig-name descname"><span class="pre">predict</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">idx</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">int</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.predict"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a 
+<dt class="sig sig-object py" id="super_gradients.training.Trainer.predict">
 
+<span class="sig-name descname"><span class="pre">predict</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">idx</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">int</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#super_gradients.training.Trainer.predict" title="Permalink to this definition"></a></dt>
 
 <dd><p>returns the predictions and label of the current inputs</p>
 
 <dd><p>returns the predictions and label of the current inputs</p>
 
 </dd></dl>
 
 </dd></dl>
 
 
 
@@ -175,39 +684,62 @@
 
 </dd></dl>
 
 </dd></dl>
 
 
 
 <dl class="py method">
 
 <dl class="py method">
 
-<dt class="sig sig-object py" id="super_gradients.training.SgModel.connect_dataset_interface">
 
-<span class="sig-name descname"><span class="pre">connect_dataset_interface</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">dataset_interface</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><a class="reference internal" href="super_gradients.training.datasets.dataset_interfaces.html#super_gradients.training.datasets.dataset_interfaces.dataset_interface.DatasetInterface" title="super_gradients.training.datasets.d
 
-<dd><dl class="field-list simple">
 
+<dt class="sig sig-object py" id="super_gradients.training.Trainer.train_from_config">
 
+<em class="property"><span class="pre">classmethod</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">train_from_config</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">cfg</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Union</span><span class="p"><span class="pre">[</span></span><span class="pre">DictConfig</span><span class="p"><span cl
 
+<dd><p>Trains according to cfg recipe configuration.</p>
 
+<p>&#64;param cfg: The parsed DictConfig from yaml recipe files or a dictionary
 
+&#64;return: the model and the output of trainer.train(…) (i.e results tuple)</p>
 
+</dd></dl>
 
+
 
+<dl class="py method">
 
+<dt class="sig sig-object py" id="super_gradients.training.Trainer.resume_experiment">
 
+<em class="property"><span class="pre">classmethod</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">resume_experiment</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">experiment_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">ckpt_root_dir</span></span><
 
+<dd><p>Resume a training that was run using our recipes.</p>
 
+<dl class="field-list simple">
 
 <dt class="field-odd">Parameters</dt>
 
 <dt class="field-odd">Parameters</dt>
 
 <dd class="field-odd"><ul class="simple">
 
 <dd class="field-odd"><ul class="simple">
 
-<li><p><strong>dataset_interface</strong> – DatasetInterface object</p></li>
 
-<li><p><strong>data_loader_num_workers</strong> – The number of threads to initialize the Data Loaders with
 
-The dataset to be connected</p></li>
 
+<li><p><strong>experiment_name</strong> – Name of the experiment to resume</p></li>
 
+<li><p><strong>ckpt_root_dir</strong> – Directory including the checkpoints</p></li>
 
 </ul>
 
 </ul>
 
 </dd>
 
 </dd>
 
 </dl>
 
 </dl>
 
 </dd></dl>
 
 </dd></dl>
 
 
 
 <dl class="py method">
 
 <dl class="py method">
 
-<dt class="sig sig-object py" id="super_gradients.training.SgModel.build_model">
 
-<span class="sig-name descname"><span class="pre">build_model</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">architecture</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">Union</span><span class="p"><span class="pre">[</span></span><span class="pre">str</span><span class="p"><span class="pre">,</span> </span><span class="pre">torch.nn.modules.module.Module</span><span class="p"><span class="pre
 
-<dd><dl class="field-list simple">
 
+<dt class="sig sig-object py" id="super_gradients.training.Trainer.evaluate_from_recipe">
 
+<em class="property"><span class="pre">classmethod</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">evaluate_from_recipe</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">cfg</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">DictConfig</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">
 
+<dd><p>Evaluate according to a cfg recipe configuration.</p>
 
+<dl class="simple">
 
+<dt>Note:   This script does NOT run training, only validation.</dt><dd><p>Please make sure that the config refers to a PRETRAINED MODEL either from one of your checkpoint or from pretrained weights from model zoo.</p>
 
+</dd>
 
+</dl>
 
+<dl class="field-list simple">
 
+<dt class="field-odd">Parameters</dt>
 
+<dd class="field-odd"><p><strong>cfg</strong> – The parsed DictConfig from yaml recipe files or a dictionary</p>
 
+</dd>
 
+</dl>
 
+</dd></dl>
 
+
 
+<dl class="py method">
 
+<dt class="sig sig-object py" id="super_gradients.training.Trainer.evaluate_checkpoint">
 
+<em class="property"><span class="pre">classmethod</span><span class="w"> </span></em><span class="sig-name descname"><span class="pre">evaluate_checkpoint</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">experiment_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">ckpt_name</span></span><sp
 
+<dd><p>Evaluate a checkpoint resulting from one of your previous experiment, using the same parameters (dataset, valid_metrics,…)
 
+as used during the training of the experiment</p>
 
+<div class="admonition note">
 
+<p class="admonition-title">Note</p>
 
+<p>The parameters will be unchanged even if the recipe used for that experiment was changed since then.
 
+This is to ensure that validation of the experiment will remain exactly the same as during training.</p>
 
+</div>
 
+<dl class="simple">
 
+<dt>Example, evaluate the checkpoint “average_model.pth” from experiment “my_experiment_name”:</dt><dd><p>&gt;&gt; evaluate_checkpoint(experiment_name=”my_experiment_name”, ckpt_name=”average_model.pth”)</p>
 
+</dd>
 
+</dl>
 
+<dl class="field-list simple">
 
 <dt class="field-odd">Parameters</dt>
 
 <dt class="field-odd">Parameters</dt>
 
 <dd class="field-odd"><ul class="simple">
 
 <dd class="field-odd"><ul class="simple">
 
-<li><p><strong>architecture</strong> – Defines the network’s architecture from models/ALL_ARCHITECTURES</p></li>
 
-<li><p><strong>arch_params</strong> – Architecture H.P. e.g.: block, num_blocks, num_classes, etc.</p></li>
 
-<li><p><strong>checkpoint_params</strong> – <p>Dictionary like object with the following key:values:</p>
 
-<p>load_checkpoint:            Load a pre-trained checkpoint
 
-strict_load:                See StrictLoad class documentation for details.
 
-source_ckpt_folder_name:    folder name to load the checkpoint from (self.experiment_name if none is given)
 
-load_weights_only:          loads only the weight from the checkpoint and zeroize the training params
 
-load_backbone:              loads the provided checkpoint to self.net.backbone instead of self.net
 
-external_checkpoint_path:   The path to the external checkpoint to be loaded. Can be absolute or relative</p>
 
-<blockquote>
 
-<div><p>(ie: path/to/checkpoint.pth). If provided, will automatically attempt to
 
-load the checkpoint even if the load_checkpoint flag is not provided.</p>
 
-</div></blockquote>
 
-</p></li>
 
+<li><p><strong>experiment_name</strong> – Name of the experiment to validate</p></li>
 
+<li><p><strong>ckpt_name</strong> – Name of the checkpoint to test (“ckpt_latest.pth”, “average_model.pth” or “ckpt_best.pth” for instance)</p></li>
 
+<li><p><strong>ckpt_root_dir</strong> – Directory including the checkpoints</p></li>
 
 </ul>
 
 </ul>
 
 </dd>
 
 </dd>
 
 </dl>
 
 </dl>
 
@@ -215,15 +747,49 @@ load the checkpoint even if the load_checkpoint flag is not provided.</p>
 
 
 
 <dl class="py method">
 
 <dl class="py method">
 
 <dt class="sig sig-object py" id="id0">
 
 <dt class="sig sig-object py" id="id0">
 
-<span class="sig-name descname"><span class="pre">train</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">training_params</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">dict</span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">{}</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/
 
+<span class="sig-name descname"><span class="pre">train</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">model</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">Module</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">training_params</span></span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pr
 
 <dd><p>train - Trains the Model</p>
 
 <dd><p>train - Trains the Model</p>
 
 <dl>
 
 <dl>
 
 <dt>IMPORTANT NOTE: Additional batch parameters can be added as a third item (optional) if a tuple is returned by</dt><dd><p>the data loaders, as dictionary. The phase context will hold the additional items, under an attribute with
 
 <dt>IMPORTANT NOTE: Additional batch parameters can be added as a third item (optional) if a tuple is returned by</dt><dd><p>the data loaders, as dictionary. The phase context will hold the additional items, under an attribute with
 
 the same name as the key in this dictionary. Then such items can be accessed through phase callbacks.</p>
 
 the same name as the key in this dictionary. Then such items can be accessed through phase callbacks.</p>
 
 <blockquote>
 
 <blockquote>
 
 <div><dl class="field-list">
 
 <div><dl class="field-list">
 
+<dt class="field-odd">param additional_configs_to_log</dt>
 
+<dd class="field-odd"><p>Dict, dictionary containing configs that will be added to the training’s
 
+sg_logger. Format should be {“Config_title_1”: {…}, “Config_title_2”:{..}}.</p>
 
+</dd>
 
+<dt class="field-even">param model</dt>
 
+<dd class="field-even"><p>torch.nn.Module, model to train.</p>
 
+</dd>
 
+<dt class="field-odd">param train_loader</dt>
 
+<dd class="field-odd"><p>Dataloader for train set.</p>
 
+</dd>
 
+<dt class="field-even">param valid_loader</dt>
 
+<dd class="field-even"><p>Dataloader for validation.</p>
 
+</dd>
 
 <dt class="field-odd">param training_params</dt>
 
 <dt class="field-odd">param training_params</dt>
 
 <dd class="field-odd"><ul>
 
 <dd class="field-odd"><ul>
 
+<li><p><cite>resume</cite> : bool (default=False)</p>
 
+<blockquote>
 
+<div><dl class="simple">
 
+<dt>Whether to continue training from ckpt with the same experiment name</dt><dd><p>(i.e resume from CKPT_ROOT_DIR/EXPERIMENT_NAME/CKPT_NAME)</p>
 
+</dd>
 
+</dl>
 
+</div></blockquote>
 
+</li>
 
+<li><p><cite>ckpt_name</cite> : str (default=ckpt_latest.pth)</p>
 
+<blockquote>
 
+<div><dl class="simple">
 
+<dt>The checkpoint (.pth file) filename in CKPT_ROOT_DIR/EXPERIMENT_NAME/ to use when resume=True and</dt><dd><p>resume_path=None</p>
 
+</dd>
 
+</dl>
 
+</div></blockquote>
 
+</li>
 
+<li><p><cite>resume_path</cite>: str (default=None)</p>
 
+<blockquote>
 
+<div><p>Explicit checkpoint path (.pth file) to use to resume training.</p>
 
+</div></blockquote>
 
+</li>
 
 <li><p><cite>max_epochs</cite> : int</p>
 
 <li><p><cite>max_epochs</cite> : int</p>
 
 <blockquote>
 
 <blockquote>
 
 <div><p>Number of epochs to run training.</p>
 
 <div><p>Number of epochs to run training.</p>
 
@@ -274,6 +840,7 @@ in each epoch iteration <cite>self.lr = self.initial_lr * pow((1.0 - (current_it
 
 </li>
 
 </li>
 
 <li><p><cite>loss</cite> : Union[nn.module, str]</p>
 
 <li><p><cite>loss</cite> : Union[nn.module, str]</p>
 
 <blockquote>
 
 <blockquote>
 
+<div><blockquote>
 
 <div><p>Loss function for training.
 
 <div><p>Loss function for training.
 
 One of SuperGradient’s built in options:</p>
 
 One of SuperGradient’s built in options:</p>
 
 <blockquote>
 
 <blockquote>
 
@@ -292,12 +859,78 @@ shape (n_items), of values computed during the forward pass which we desire to l
 
 entire epoch. For example- the loss itself should always be logged. Another example is a scenario
 
 entire epoch. For example- the loss itself should always be logged. Another example is a scenario
 
 where the computed loss is the sum of a few components we would like to log- these entries in
 
 where the computed loss is the sum of a few components we would like to log- these entries in
 
 loss_items).</p>
 
 loss_items).</p>
 
-<p>When training, set the loss_logging_items_names parameter in train_params to be a list of
 
-strings, of length n_items who’s ith element is the name of the ith entry in loss_items. Then
 
-each item will be logged, rendered on tensorboard and “watched” (i.e saving model checkpoints
 
-according to it).</p>
 
+<p>IMPORTANT:When dealing with external loss classes, to logg/monitor the loss_items as described
 
+above by specific string name:</p>
 
+<dl>
 
+<dt>Set a “component_names” property in the loss class, whos instance is passed through train_params,</dt><dd><p>to be a list of strings, of length n_items who’s ith element is the name of the ith entry in loss_items.
 
+Then each item will be logged, rendered on tensorboard and “watched” (i.e saving model checkpoints
 
+according to it) under &lt;LOSS_CLASS.__name__&gt;”/”&lt;COMPONENT_NAME&gt;. If a single item is returned rather then a
 
+tuple, it would be logged under &lt;LOSS_CLASS.__name__&gt;. When there is no such attributed, the items
 
+will be named &lt;LOSS_CLASS.__name__&gt;”/”<a href="#id6"><span class="problematic" id="id7">Loss_</span></a>”&lt;IDX&gt; according to the length of loss_items</p>
 
+</dd>
 
+<dt>For example:</dt><dd><dl>
 
+<dt>class MyLoss(_Loss):</dt><dd><p>…
 
+def forward(self, inputs, targets):</p>
 
+<blockquote>
 
+<div><p>…
 
+total_loss = comp1 + comp2
 
+loss_items = torch.cat((total_loss.unsqueeze(0),comp1.unsqueeze(0), comp2.unsqueeze(0)).detach()
 
+return total_loss, loss_items</p>
 
+</div></blockquote>
 
+<p>…
 
+&#64;property
 
+def component_names(self):</p>
 
+<blockquote>
 
+<div><p>return [“total_loss”, “my_1st_component”, “my_2nd_component”]</p>
 
+</div></blockquote>
 
+</dd>
 
+</dl>
 
+</dd>
 
+<dt>Trainer.train(…</dt><dd><blockquote>
 
+<div><dl class="simple">
 
+<dt>train_params={“loss”:MyLoss(),</dt><dd><p>…
 
+“metric_to_watch”: “MyLoss/my_1st_component”}</p>
 
+</dd>
 
+</dl>
 
+</div></blockquote>
 
+<dl class="simple">
 
+<dt>This will write to log and monitor MyLoss/total_loss, MyLoss/my_1st_component,</dt><dd><p>MyLoss/my_2nd_component.</p>
 
+</dd>
 
+</dl>
 
+</dd>
 
+</dl>
 
+</div></blockquote>
 
+<dl>
 
+<dt>For example:</dt><dd><blockquote>
 
+<div><dl>
 
+<dt>class MyLoss2(_Loss):</dt><dd><p>…
 
+def forward(self, inputs, targets):</p>
 
+<blockquote>
 
+<div><p>…
 
+total_loss = comp1 + comp2
 
+loss_items = torch.cat((total_loss.unsqueeze(0),comp1.unsqueeze(0), comp2.unsqueeze(0)).detach()
 
+return total_loss, loss_items</p>
 
+</div></blockquote>
 
+<p>…</p>
 
+</dd>
 
+</dl>
 
+</div></blockquote>
 
+<dl>
 
+<dt>Trainer.train(…</dt><dd><blockquote>
 
+<div><dl class="simple">
 
+<dt>train_params={“loss”:MyLoss(),</dt><dd><p>…
 
+“metric_to_watch”: “MyLoss2/loss_0”}</p>
 
+</dd>
 
+</dl>
 
+</div></blockquote>
 
+<p>This will write to log and monitor MyLoss2/loss_0, MyLoss2/loss_1, MyLoss2/loss_2
 
+as they have been named by their positional index in loss_items.</p>
 
+</dd>
 
+</dl>
 
 <p>Since running logs will save the loss_items in some internal state, it is recommended that
 
 <p>Since running logs will save the loss_items in some internal state, it is recommended that
 
 loss_items are detached from their computational graph for memory efficiency.</p>
 
 loss_items are detached from their computational graph for memory efficiency.</p>
 
+</dd>
 
+</dl>
 
 </div></blockquote>
 
 </div></blockquote>
 
 </li>
 
 </li>
 
 <li><p><cite>optimizer</cite> : Union[str, torch.optim.Optimizer]</p>
 
 <li><p><cite>optimizer</cite> : Union[str, torch.optim.Optimizer]</p>
 
@@ -346,7 +979,7 @@ of the following:</p>
 
 <p>a “metric_name” if some metric in valid_metrics_list has an attribute component_names which
 
 <p>a “metric_name” if some metric in valid_metrics_list has an attribute component_names which
 
 is a list referring to the names of each entry in the output metric (torch tensor of size n)</p>
 
 is a list referring to the names of each entry in the output metric (torch tensor of size n)</p>
 
 <p>one of “loss_logging_items_names” i.e which will correspond to an item returned during the
 
 <p>one of “loss_logging_items_names” i.e which will correspond to an item returned during the
 
-loss function’s forward pass.</p>
 
+loss function’s forward pass (see loss docs abov).</p>
 
 </div></blockquote>
 
 </div></blockquote>
 
 <p>At the end of each epoch, if a new best metric_to_watch value is achieved, the models checkpoint
 
 <p>At the end of each epoch, if a new best metric_to_watch value is achieved, the models checkpoint
 
 is saved in YOUR_PYTHON_PATH/checkpoints/ckpt_best.pth</p>
 
 is saved in YOUR_PYTHON_PATH/checkpoints/ckpt_best.pth</p>
 
@@ -456,14 +1089,6 @@ will be added to the tensorboard along with some sample images from the dataset.
 
 detection datasets are supported for analysis.</p>
 
 detection datasets are supported for analysis.</p>
 
 </div></blockquote>
 
 </div></blockquote>
 
 </li>
 
 </li>
 
-<li><p><cite>save_full_train_log</cite> : bool (default=False)</p>
 
-<blockquote>
 
-<div><dl class="simple">
 
-<dt>When set, a full log (of all super_gradients modules, including uncaught exceptions from any other</dt><dd><p>module) of the training will be saved in the checkpoint directory under full_train_log.log</p>
 
-</dd>
 
-</dl>
 
-</div></blockquote>
 
-</li>
 
 <li><p><cite>sg_logger</cite> : Union[AbstractSGLogger, str] (defauls=base_sg_logger)</p>
 
 <li><p><cite>sg_logger</cite> : Union[AbstractSGLogger, str] (defauls=base_sg_logger)</p>
 
 <blockquote>
 
 <blockquote>
 
 <div><p>Define the SGLogger object for this training process. The SGLogger handles all disk writes, logs, TensorBoard, remote logging
 
 <div><p>Define the SGLogger object for this training process. The SGLogger handles all disk writes, logs, TensorBoard, remote logging
 
@@ -514,7 +1139,7 @@ or support remote storage.</p>
 
 </dl>
 
 </dl>
 
 <p>num_calib_batches: int, number of batches to collect the statistics from.</p>
 
 <p>num_calib_batches: int, number of batches to collect the statistics from.</p>
 
 <dl class="simple">
 
 <dl class="simple">
 
-<dt>percentile: float, percentile value to use when SgModel,quant_modules_calib_method=’percentile’.</dt><dd><p>Discarded when other methods are used (Default=99.99).</p>
 
+<dt>percentile: float, percentile value to use when Trainer,quant_modules_calib_method=’percentile’.</dt><dd><p>Discarded when other methods are used (Default=99.99).</p>
 
 </dd>
 
 </dd>
 
 </dl>
 
 </dl>
 
 </div></blockquote>
 
 </div></blockquote>
 
@@ -532,96 +1157,63 @@ or support remote storage.</p>
 
 </dl>
 
 </dl>
 
 </dd></dl>
 
 </dd></dl>
 
 
 
-<dl class="py method">
 
-<dt class="sig sig-object py" id="id3">
 
-<span class="sig-name descname"><span class="pre">predict</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">inputs</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">targets</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">half</span></span><span class="o"><span class="pre">=</span>
 
-<dd><p>A fast predictor for a batch of inputs
 
-:param inputs: torch.tensor or numpy.array</p>
 
-<blockquote>
 
-<div><p>a batch of inputs</p>
 
-</div></blockquote>
 
-<dl class="field-list simple">
 
-<dt class="field-odd">Parameters</dt>
 
-<dd class="field-odd"><ul class="simple">
 
-<li><p><strong>targets</strong> – torch.tensor()
 
-corresponding labels - if non are given - accuracy will not be computed</p></li>
 
-<li><p><strong>verbose</strong> – bool
 
-print the results to screen</p></li>
 
-<li><p><strong>normalize</strong> – bool
 
-If true, normalizes the tensor according to the dataloader’s normalization values</p></li>
 
-<li><p><strong>half</strong> – Performs half precision evaluation</p></li>
 
-<li><p><strong>move_outputs_to_cpu</strong> – Moves the results from the GPU to the CPU</p></li>
 
-</ul>
 
-</dd>
 
-<dt class="field-even">Returns</dt>
 
-<dd class="field-even"><p>outputs, acc, net_time, gross_time
Discard
Prev
Next
Tip!

Press p or to see the previous file or, n or to see the next file