deploy: SimplexLab/TorchJD@b4a6826

Author: ValerianRey

latest/docs/aggregation/aligned_mtl/index.html

@@ -311,7 +311,8 @@ <h1>Aligned-MTL<a class="headerlink" href="#aligned-mtl" title="Link to this hea
 </dl>
 <div class="admonition note">
 <p class="admonition-title">Note</p>
-<p>This implementation was adapted from the <a class="reference external" href="https://github.com/SamsungLabs/MTL/tree/master/code/optim/aligned">official implementation</a>.</p>
+<p>This implementation was adapted from the official implementation of SamsungLabs/MTL,
+which is not available anymore at the time of writing.</p>
 </div>
 </dd></dl>
 

latest/docs/aggregation/mgda/index.html

@@ -295,10 +295,11 @@
 <h1>MGDA<a class="headerlink" href="#mgda" title="Link to this heading">¶</a></h1>
 <dl class="py class">
 <dt class="sig sig-object py" id="torchjd.aggregation.MGDA">
-<span class="property"><span class="k"><span class="pre">class</span></span><span class="w"> </span></span><span class="sig-prename descclassname"><span class="pre">torchjd.aggregation.</span></span><span class="sig-name descname"><span class="pre">MGDA</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">epsilon</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">0.001</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">max_iters</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">100</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/TorchJD/torchjd/blob/main/src/torchjd/aggregation/_mgda.py#L10-L28"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#torchjd.aggregation.MGDA" title="Link to this definition">¶</a></dt>
+<span class="property"><span class="k"><span class="pre">class</span></span><span class="w"> </span></span><span class="sig-prename descclassname"><span class="pre">torchjd.aggregation.</span></span><span class="sig-name descname"><span class="pre">MGDA</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">epsilon</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">0.001</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">max_iters</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">100</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/TorchJD/torchjd/blob/main/src/torchjd/aggregation/_mgda.py#L10-L29"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#torchjd.aggregation.MGDA" title="Link to this definition">¶</a></dt>
 <dd><p><a class="reference internal" href="../#torchjd.aggregation.Aggregator" title="torchjd.aggregation._aggregator_bases.Aggregator"><code class="xref py py-class docutils literal notranslate"><span class="pre">Aggregator</span></code></a> performing the gradient aggregation
-step of <a class="reference external" href="https://www.sciencedirect.com/science/article/pii/S1631073X12000738">Multiple-gradient descent algorithm (MGDA) for multiobjective optimization</a>. The implementation is
-based on Algorithm 2 of <a class="reference external" href="https://proceedings.neurips.cc/paper_files/paper/2018/file/432aca3a1e345e339f35a30c8f65edce-Paper.pdf">Multi-Task Learning as Multi-Objective Optimization</a>.</p>
+step of <a class="reference external" href="https://comptes-rendus.academie-sciences.fr/mathematique/articles/10.1016/j.crma.2012.03.014/">Multiple-gradient descent algorithm (MGDA) for multiobjective optimization</a>.
+The implementation is based on Algorithm 2 of <a class="reference external" href="https://proceedings.neurips.cc/paper_files/paper/2018/file/432aca3a1e345e339f35a30c8f65edce-Paper.pdf">Multi-Task Learning as Multi-Objective
+Optimization</a>.</p>
 <dl class="field-list simple">
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
@@ -311,7 +312,7 @@ <h1>MGDA<a class="headerlink" href="#mgda" title="Link to this heading">¶</a></
 
 <dl class="py class">
 <dt class="sig sig-object py" id="torchjd.aggregation.MGDAWeighting">
-<span class="property"><span class="k"><span class="pre">class</span></span><span class="w"> </span></span><span class="sig-prename descclassname"><span class="pre">torchjd.aggregation.</span></span><span class="sig-name descname"><span class="pre">MGDAWeighting</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">epsilon</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">0.001</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">max_iters</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">100</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/TorchJD/torchjd/blob/main/src/torchjd/aggregation/_mgda.py#L31-L71"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#torchjd.aggregation.MGDAWeighting" title="Link to this definition">¶</a></dt>
+<span class="property"><span class="k"><span class="pre">class</span></span><span class="w"> </span></span><span class="sig-prename descclassname"><span class="pre">torchjd.aggregation.</span></span><span class="sig-name descname"><span class="pre">MGDAWeighting</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">epsilon</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">0.001</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">max_iters</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">100</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/TorchJD/torchjd/blob/main/src/torchjd/aggregation/_mgda.py#L32-L72"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#torchjd.aggregation.MGDAWeighting" title="Link to this definition">¶</a></dt>
 <dd><p><a class="reference internal" href="../#torchjd.aggregation.Weighting" title="torchjd.aggregation._weighting_bases.Weighting"><code class="xref py py-class docutils literal notranslate"><span class="pre">Weighting</span></code></a> giving the weights of
 <a class="reference internal" href="#torchjd.aggregation.MGDA" title="torchjd.aggregation.MGDA"><code class="xref py py-class docutils literal notranslate"><span class="pre">MGDA</span></code></a>.</p>
 <dl class="field-list simple">

latest/examples/lightning_integration/index.html

@@ -293,12 +293,12 @@
         <article role="main" id="furo-main-content">
           <section id="pytorch-lightning-integration">
 <h1>PyTorch Lightning Integration<a class="headerlink" href="#pytorch-lightning-integration" title="Link to this heading">¶</a></h1>
-<p>To use Jacobian descent with TorchJD in a <a class="reference external" href="https://lightning.ai/docs/pytorch/stable/api/lightning.pytorch.core.LightningModule.html#lightning.pytorch.core.LightningModule" title="(in PyTorch Lightning v2.6.0)"><code class="xref py py-class docutils literal notranslate"><span class="pre">LightningModule</span></code></a>, you need
+<p>To use Jacobian descent with TorchJD in a <a class="reference external" href="https://lightning.ai/docs/pytorch/stable/api/lightning.pytorch.core.LightningModule.html#lightning.pytorch.core.LightningModule" title="(in PyTorch Lightning v2.6.1)"><code class="xref py py-class docutils literal notranslate"><span class="pre">LightningModule</span></code></a>, you need
 to turn off automatic optimization by setting <code class="docutils literal notranslate"><span class="pre">automatic_optimization</span></code> to <code class="docutils literal notranslate"><span class="pre">False</span></code> and to
 customize the <code class="docutils literal notranslate"><span class="pre">training_step</span></code> method to make it call the appropriate TorchJD method
 (<a class="reference internal" href="../../docs/autojac/backward/"><span class="doc">backward</span></a> or <a class="reference internal" href="../../docs/autojac/mtl_backward/"><span class="doc">mtl_backward</span></a>).</p>
 <p>The following code example demonstrates a basic multi-task learning setup using a
-<a class="reference external" href="https://lightning.ai/docs/pytorch/stable/api/lightning.pytorch.core.LightningModule.html#lightning.pytorch.core.LightningModule" title="(in PyTorch Lightning v2.6.0)"><code class="xref py py-class docutils literal notranslate"><span class="pre">LightningModule</span></code></a> that will call <a class="reference internal" href="../../docs/autojac/mtl_backward/"><span class="doc">mtl_backward</span></a> at each training iteration.</p>
+<a class="reference external" href="https://lightning.ai/docs/pytorch/stable/api/lightning.pytorch.core.LightningModule.html#lightning.pytorch.core.LightningModule" title="(in PyTorch Lightning v2.6.1)"><code class="xref py py-class docutils literal notranslate"><span class="pre">LightningModule</span></code></a> that will call <a class="reference internal" href="../../docs/autojac/mtl_backward/"><span class="doc">mtl_backward</span></a> at each training iteration.</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span><span class="w"> </span><span class="nn">torch</span>
 <span class="kn">from</span><span class="w"> </span><span class="nn">lightning</span><span class="w"> </span><span class="kn">import</span> <span class="n">LightningModule</span><span class="p">,</span> <span class="n">Trainer</span>
 <span class="kn">from</span><span class="w"> </span><span class="nn">lightning.pytorch.utilities.types</span><span class="w"> </span><span class="kn">import</span> <span class="n">OptimizerLRScheduler</span>