auto-generating sphinx docs

Author: pytorchbot

main/_modules/torchtune/training/_profiler.html

@@ -475,6 +475,7 @@ <h1>Source code for torchtune.training._profiler</h1><div class="highlight"><pre
 <span class="n">DEFAULT_PROFILER_ACTIVITIES</span> <span class="o">=</span> <span class="p">{</span>
     <span class="n">torch</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">ProfilerActivity</span><span class="o">.</span><span class="n">CPU</span><span class="p">,</span>
     <span class="n">torch</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">ProfilerActivity</span><span class="o">.</span><span class="n">CUDA</span><span class="p">,</span>
+    <span class="n">torch</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">ProfilerActivity</span><span class="o">.</span><span class="n">XPU</span><span class="p">,</span>
 <span class="p">}</span>
 
 <span class="n">DEFAULT_SCHEDULE</span><span class="p">:</span> <span class="nb">dict</span> <span class="o">=</span> <span class="p">{</span>
@@ -559,7 +560,7 @@ <h1>Source code for torchtune.training._profiler</h1><div class="highlight"><pre
         <span class="n">log</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">&quot;Finished dumping traces in </span><span class="si">{</span><span class="n">time</span><span class="o">.</span><span class="n">monotonic</span><span class="p">()</span><span class="w"> </span><span class="o">-</span><span class="w"> </span><span class="n">begin</span><span class="si">:</span><span class="s2">.2f</span><span class="si">}</span><span class="s2"> seconds&quot;</span><span class="p">)</span>
 
     <span class="c1"># Memory timeline sometimes fails to export</span>
-    <span class="k">if</span> <span class="n">prof</span><span class="o">.</span><span class="n">profile_memory</span><span class="p">:</span>
+    <span class="k">if</span> <span class="n">prof</span><span class="o">.</span><span class="n">profile_memory</span> <span class="ow">and</span> <span class="n">torch</span><span class="o">.</span><span class="n">cuda</span><span class="o">.</span><span class="n">is_available</span><span class="p">():</span>
         <span class="k">if</span> <span class="n">rank</span> <span class="o">==</span> <span class="mi">0</span><span class="p">:</span>
             <span class="k">try</span><span class="p">:</span>
                 <span class="n">prof</span><span class="o">.</span><span class="n">export_memory_timeline</span><span class="p">(</span>
@@ -633,6 +634,7 @@ <h1>Source code for torchtune.training._profiler</h1><div class="highlight"><pre
     <span class="n">enabled</span><span class="p">:</span> <span class="nb">bool</span> <span class="o">=</span> <span class="kc">False</span><span class="p">,</span>
     <span class="n">cpu</span><span class="p">:</span> <span class="nb">bool</span> <span class="o">=</span> <span class="kc">True</span><span class="p">,</span>
     <span class="n">cuda</span><span class="p">:</span> <span class="nb">bool</span> <span class="o">=</span> <span class="kc">True</span><span class="p">,</span>
+    <span class="n">xpu</span><span class="p">:</span> <span class="nb">bool</span> <span class="o">=</span> <span class="kc">True</span><span class="p">,</span>
     <span class="n">profile_memory</span><span class="p">:</span> <span class="nb">bool</span> <span class="o">=</span> <span class="n">DEFAULT_TRACE_OPTS</span><span class="p">[</span><span class="s2">&quot;profile_memory&quot;</span><span class="p">],</span>
     <span class="n">with_stack</span><span class="p">:</span> <span class="nb">bool</span> <span class="o">=</span> <span class="n">DEFAULT_TRACE_OPTS</span><span class="p">[</span><span class="s2">&quot;with_stack&quot;</span><span class="p">],</span>
     <span class="n">record_shapes</span><span class="p">:</span> <span class="nb">bool</span> <span class="o">=</span> <span class="n">DEFAULT_TRACE_OPTS</span><span class="p">[</span><span class="s2">&quot;record_shapes&quot;</span><span class="p">],</span>
@@ -700,6 +702,7 @@ <h1>Source code for torchtune.training._profiler</h1><div class="highlight"><pre
 <span class="sd">        enabled (bool): Enable pytorch profiler. Default is False.</span>
 <span class="sd">        cpu (bool): Enable cpu profiling. Default is True.</span>
 <span class="sd">        cuda (bool): Enable cuda profiling. Default is True.</span>
+<span class="sd">        xpu (bool): Enable xpu profiling. Default is True.</span>
 <span class="sd">        profile_memory (bool): Profile memory usage. Default is False.</span>
 <span class="sd">        with_stack (bool): Profile stack. Default is False.</span>
 <span class="sd">        record_shapes (bool): Record shapes. Default is True.</span>
@@ -724,10 +727,12 @@ <h1>Source code for torchtune.training._profiler</h1><div class="highlight"><pre
         <span class="n">activities</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">ProfilerActivity</span><span class="o">.</span><span class="n">CPU</span><span class="p">)</span>
     <span class="k">if</span> <span class="n">cuda</span><span class="p">:</span>
         <span class="n">activities</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">ProfilerActivity</span><span class="o">.</span><span class="n">CUDA</span><span class="p">)</span>
+    <span class="k">if</span> <span class="n">xpu</span><span class="p">:</span>
+        <span class="n">activities</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">ProfilerActivity</span><span class="o">.</span><span class="n">XPU</span><span class="p">)</span>
     <span class="k">if</span> <span class="nb">len</span><span class="p">(</span><span class="n">activities</span><span class="p">)</span> <span class="o">==</span> <span class="mi">0</span><span class="p">:</span>
         <span class="n">_warn</span><span class="p">(</span><span class="s2">&quot;No activities specified, defaulting to CPU + CUDA&quot;</span><span class="p">)</span>
         <span class="n">activities</span> <span class="o">=</span> <span class="n">DEFAULT_PROFILER_ACTIVITIES</span>
-        <span class="n">cpu</span> <span class="o">=</span> <span class="n">cuda</span> <span class="o">=</span> <span class="kc">True</span>
+        <span class="n">cpu</span> <span class="o">=</span> <span class="n">cuda</span> <span class="o">=</span> <span class="n">xpu</span> <span class="o">=</span> <span class="kc">True</span>
 
     <span class="c1"># Check for schedule</span>
     <span class="c1"># 1) If no schedule is provided, set to DEFAULT_SCHEDULE</span>
@@ -820,6 +825,7 @@ <h1>Source code for torchtune.training._profiler</h1><div class="highlight"><pre
             <span class="s2">&quot;output_dir&quot;</span><span class="p">:</span> <span class="n">output_dir</span><span class="p">,</span>
             <span class="s2">&quot;cpu&quot;</span><span class="p">:</span> <span class="n">cpu</span><span class="p">,</span>
             <span class="s2">&quot;cuda&quot;</span><span class="p">:</span> <span class="n">cuda</span><span class="p">,</span>
+            <span class="s2">&quot;xpu&quot;</span><span class="p">:</span> <span class="n">xpu</span><span class="p">,</span>
             <span class="s2">&quot;profile_memory&quot;</span><span class="p">:</span> <span class="n">profile_memory</span><span class="p">,</span>
             <span class="s2">&quot;with_stack&quot;</span><span class="p">:</span> <span class="n">with_stack</span><span class="p">,</span>
             <span class="s2">&quot;record_shapes&quot;</span><span class="p">:</span> <span class="n">record_shapes</span><span class="p">,</span>