Add changes for a48abc1

Author: actions-user

latest/_sources/optimization.rst.txt

@@ -46,6 +46,12 @@ cache files, serving a value from the cache for the first time in the run also c
 Only unique function evaluations are counted, so the second time a parameter configuration is selected by the strategy it is served from the 
 cache, but not counted as a unique function evaluation.
 
+All optimization algorithms, except for brute_force, random_sample, and bayes_opt, allow the user to specify an initial guess or 
+starting point for the optimization, called ``x0``. This can be passed to the strategy using the ``strategy_options=`` dictionary with ``"x0"`` as key and
+a list of values for each parameter in tune_params to note the starting point. For example, for a kernel that has parameters ``block_size_x`` (64, 128, 256)
+and ``tile_size_x`` (1,2,3), one could pass ``strategy_options=dict(x0=[128,2])`` to ``tune_kernel()`` to make sure the strategy starts from
+the configuration with ``block_size_x=128, tile_size_x=2``. The order in the ``x0`` list should match the order in the tunable parameters dictionary.
+
 Below all the strategies are listed with their strategy-specific options that can be passed in a dictionary to the ``strategy_options=`` argument
 of ``tune_kernel()``.
 

latest/design.html

@@ -270,7 +270,7 @@ <h2>Strategies<a class="headerlink" href="#strategies" title="Link to this headi
 <span id="kernel-tuner-strategies-common"></span><h3>kernel_tuner.strategies.common<a class="headerlink" href="#module-kernel_tuner.strategies.common" title="Link to this heading">¶</a></h3>
 <dl class="py function">
 <dt class="sig sig-object py" id="kernel_tuner.strategies.common.get_options">
-<span class="sig-prename descclassname"><span class="pre">kernel_tuner.strategies.common.</span></span><span class="sig-name descname"><span class="pre">get_options</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">strategy_options</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">options</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#kernel_tuner.strategies.common.get_options" title="Link to this definition">¶</a></dt>
+<span class="sig-prename descclassname"><span class="pre">kernel_tuner.strategies.common.</span></span><span class="sig-name descname"><span class="pre">get_options</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">strategy_options</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">options</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">unsupported</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#kernel_tuner.strategies.common.get_options" title="Link to this definition">¶</a></dt>
 <dd><p>Get the strategy-specific options or their defaults from user-supplied strategy_options.</p>
 </dd></dl>
 

latest/optimization.html

@@ -224,6 +224,11 @@
 cache files, serving a value from the cache for the first time in the run also counts as a function evaluation for the strategy.
 Only unique function evaluations are counted, so the second time a parameter configuration is selected by the strategy it is served from the
 cache, but not counted as a unique function evaluation.</p>
+<p>All optimization algorithms, except for brute_force, random_sample, and bayes_opt, allow the user to specify an initial guess or
+starting point for the optimization, called <code class="docutils literal notranslate"><span class="pre">x0</span></code>. This can be passed to the strategy using the <code class="docutils literal notranslate"><span class="pre">strategy_options=</span></code> dictionary with <code class="docutils literal notranslate"><span class="pre">&quot;x0&quot;</span></code> as key and
+a list of values for each parameter in tune_params to note the starting point. For example, for a kernel that has parameters <code class="docutils literal notranslate"><span class="pre">block_size_x</span></code> (64, 128, 256)
+and <code class="docutils literal notranslate"><span class="pre">tile_size_x</span></code> (1,2,3), one could pass <code class="docutils literal notranslate"><span class="pre">strategy_options=dict(x0=[128,2])</span></code> to <code class="docutils literal notranslate"><span class="pre">tune_kernel()</span></code> to make sure the strategy starts from
+the configuration with <code class="docutils literal notranslate"><span class="pre">block_size_x=128,</span> <span class="pre">tile_size_x=2</span></code>. The order in the <code class="docutils literal notranslate"><span class="pre">x0</span></code> list should match the order in the tunable parameters dictionary.</p>
 <p>Below all the strategies are listed with their strategy-specific options that can be passed in a dictionary to the <code class="docutils literal notranslate"><span class="pre">strategy_options=</span></code> argument
 of <code class="docutils literal notranslate"><span class="pre">tune_kernel()</span></code>.</p>
 <section id="module-kernel_tuner.strategies.basinhopping">