Update docs from d068cf9

_sources/examples.md.txt

@@ -11,7 +11,6 @@
 ||red pajama|[Link](https://github.com/microsoft/Olive/tree/main/examples/red_pajama)| `CPU`: with Optimum conversion and merging and ONNX Runtime optimizations for a single optimized ONNX model
 ||bert|[Link](https://github.com/microsoft/Olive/tree/main/examples/bert)|`CPU`: with ONNX Runtime optimizations and quantization for optimized INT8 ONNX model<br>`CPU`: with ONNX Runtime optimizations and Intel® Neural Compressor quantization for optimized INT8 ONNX model<br>`CPU`: with PyTorch QAT Customized Training Loop and ONNX Runtime optimizations for optimized ONNX INT8 model<br>`GPU`: with ONNX Runtime optimizations for CUDA EP<br>`GPU`: with ONNX Runtime optimizations for TRT EP
 ||deberta|[Link](https://github.com/microsoft/Olive/tree/main/examples/deberta)|`GPU`: Optimize Azureml Registry Model with ONNX Runtime optimizations and quantization
-||dolly_v2|[Link](https://github.com/microsoft/Olive/tree/main/examples/directml/dolly_v2)|`GPU`: with Optimum conversion and merging and ONNX Runtime optimizations with DirectML EP
 ||gptj|[Link](https://github.com/microsoft/Olive/tree/main/examples/gptj)|`CPU`: with Intel® Neural Compressor static/dynamic quantization for INT8 ONNX model
 |Audio|whisper|[Link](https://github.com/microsoft/Olive/tree/main/examples/whisper)|`CPU`: with ONNX Runtime optimizations for all-in-one ONNX model in FP32<br>`CPU`: with ONNX Runtime optimizations for all-in-one ONNX model in INT8<br>`CPU`: with ONNX Runtime optimizations and Intel® Neural Compressor Dynamic Quantization for all-in-one ONNX model in INT8<br>`GPU`: with ONNX Runtime optimizations for all-in-one ONNX model in FP32<br>`GPU`: with ONNX Runtime optimizations for all-in-one ONNX model in FP16<br>`GPU`: with ONNX Runtime optimizations for all-in-one ONNX model in INT8
 ||audio spectrogram<br>transformer|[Link](https://github.com/microsoft/Olive/tree/main/examples/AST)|`CPU`: with ONNX Runtime optimizations and quantization for optimized INT8 ONNX model

api/passes.html

@@ -444,6 +444,15 @@
 <p><strong>searchable_values:</strong> None</p>
 </dd></dl>
 
+<dl class="std option">
+<dt class="sig sig-object std" id="cmdoption-arg-num_key_value_heads">
+<span class="sig-name descname"><span class="pre">num_key_value_heads</span></span><span class="sig-prename descclassname"></span><a class="headerlink" href="#cmdoption-arg-num_key_value_heads" title="Permalink to this definition">¶</a></dt>
+<dd><p>Number of key/value attention heads.</p>
+<p><strong>type:</strong> int</p>
+<p><strong>default_value:</strong> 0</p>
+<p><strong>searchable_values:</strong> None</p>
+</dd></dl>
+
 <dl class="std option">
 <dt class="sig sig-object std" id="cmdoption-arg-hidden_size">
 <span class="sig-name descname"><span class="pre">hidden_size</span></span><span class="sig-prename descclassname"></span><a class="headerlink" href="#cmdoption-arg-hidden_size" title="Permalink to this definition">¶</a></dt>

examples.html

@@ -184,56 +184,51 @@ <h1>Examples<a class="headerlink" href="#examples" title="Permalink to this head
 <td><p><code class="docutils literal notranslate"><span class="pre">GPU</span></code>: Optimize Azureml Registry Model with ONNX Runtime optimizations and quantization</p></td>
 </tr>
 <tr class="row-odd"><td><p></p></td>
-<td><p>dolly_v2</p></td>
-<td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/directml/dolly_v2">Link</a></p></td>
-<td><p><code class="docutils literal notranslate"><span class="pre">GPU</span></code>: with Optimum conversion and merging and ONNX Runtime optimizations with DirectML EP</p></td>
-</tr>
-<tr class="row-even"><td><p></p></td>
 <td><p>gptj</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/gptj">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">CPU</span></code>: with Intel® Neural Compressor static/dynamic quantization for INT8 ONNX model</p></td>
 </tr>
-<tr class="row-odd"><td><p>Audio</p></td>
+<tr class="row-even"><td><p>Audio</p></td>
 <td><p>whisper</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/whisper">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">CPU</span></code>: with ONNX Runtime optimizations for all-in-one ONNX model in FP32<br><code class="docutils literal notranslate"><span class="pre">CPU</span></code>: with ONNX Runtime optimizations for all-in-one ONNX model in INT8<br><code class="docutils literal notranslate"><span class="pre">CPU</span></code>: with ONNX Runtime optimizations and Intel® Neural Compressor Dynamic Quantization for all-in-one ONNX model in INT8<br><code class="docutils literal notranslate"><span class="pre">GPU</span></code>: with ONNX Runtime optimizations for all-in-one ONNX model in FP32<br><code class="docutils literal notranslate"><span class="pre">GPU</span></code>: with ONNX Runtime optimizations for all-in-one ONNX model in FP16<br><code class="docutils literal notranslate"><span class="pre">GPU</span></code>: with ONNX Runtime optimizations for all-in-one ONNX model in INT8</p></td>
 </tr>
-<tr class="row-even"><td><p></p></td>
+<tr class="row-odd"><td><p></p></td>
 <td><p>audio spectrogram<br>transformer</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/AST">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">CPU</span></code>: with ONNX Runtime optimizations and quantization for optimized INT8 ONNX model</p></td>
 </tr>
-<tr class="row-odd"><td><p>Vision</p></td>
+<tr class="row-even"><td><p>Vision</p></td>
 <td><p>stable diffusion <br> stable diffusion XL</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/stable_diffusion">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">GPU</span></code>: with ONNX Runtime optimization for DirectML EP<br><code class="docutils literal notranslate"><span class="pre">GPU</span></code>: with ONNX Runtime optimization for CUDA EP<br><code class="docutils literal notranslate"><span class="pre">Intel</span> <span class="pre">CPU</span></code>: with OpenVINO toolkit</p></td>
 </tr>
-<tr class="row-even"><td><p></p></td>
+<tr class="row-odd"><td><p></p></td>
 <td><p>squeezenet</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/directml/squeezenet">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">GPU</span></code>: with ONNX Runtime optimizations with DirectML EP</p></td>
 </tr>
-<tr class="row-odd"><td><p></p></td>
+<tr class="row-even"><td><p></p></td>
 <td><p>mobilenet</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/mobilenet">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">Qualcomm</span> <span class="pre">NPU</span></code>: with ONNX Runtime static QDQ quantization for ONNX Runtime QNN EP</p></td>
 </tr>
-<tr class="row-even"><td><p></p></td>
+<tr class="row-odd"><td><p></p></td>
 <td><p>resnet</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/resnet">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">CPU</span></code>: with ONNX Runtime static/dynamic Quantization for ONNX INT8 model<br><code class="docutils literal notranslate"><span class="pre">CPU</span></code>: with PyTorch QAT Default Training Loop and ONNX Runtime optimizations for ONNX INT8 model<br><code class="docutils literal notranslate"><span class="pre">CPU</span></code>: with PyTorch QAT Lightning Module and ONNX Runtime optimizations for ONNX INT8 model<br><code class="docutils literal notranslate"><span class="pre">AMD</span> <span class="pre">DPU</span></code>: with AMD Vitis-AI Quantization<br><code class="docutils literal notranslate"><span class="pre">Intel</span> <span class="pre">GPU</span></code>: with ONNX Runtime optimizations with multiple EPs</p></td>
 </tr>
-<tr class="row-odd"><td><p></p></td>
+<tr class="row-even"><td><p></p></td>
 <td><p>VGG</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/vgg">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">Qualcomm</span> <span class="pre">NPU</span></code>: with SNPE toolkit</p></td>
 </tr>
-<tr class="row-even"><td><p></p></td>
+<tr class="row-odd"><td><p></p></td>
 <td><p>inception</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/inception">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">Qualcomm</span> <span class="pre">NPU</span></code>: with SNPE toolkit</p></td>
 </tr>
-<tr class="row-odd"><td><p></p></td>
+<tr class="row-even"><td><p></p></td>
 <td><p>super resolution</p></td>
 <td><p><a class="reference external" href="https://github.com/microsoft/Olive/tree/main/examples/super_resolution">Link</a></p></td>
 <td><p><code class="docutils literal notranslate"><span class="pre">CPU</span></code>: with ONNX Runtime pre/post processing integration for a single ONNX model</p></td>

genindex.html

@@ -658,6 +658,8 @@ <h2 id="C">C</h2>
         <li><a href="api/passes.html#cmdoption-arg-num_epochs">num_epochs</a>
 </li>
         <li><a href="api/passes.html#cmdoption-arg-num_heads">num_heads</a>
+</li>
+        <li><a href="api/passes.html#cmdoption-arg-num_key_value_heads">num_key_value_heads</a>
 </li>
         <li><a href="api/search-algorithms.html#cmdoption-arg-0">num_samples</a>, <a href="api/search-algorithms.html#cmdoption-arg-num_samples">[1]</a>
 </li>
@@ -1818,6 +1820,13 @@ <h2 id="N">N</h2>
 
       <ul>
         <li><a href="api/passes.html#cmdoption-arg-num_heads">command line option</a>
+</li>
+      </ul></li>
+      <li>
+    num_key_value_heads
+
+      <ul>
+        <li><a href="api/passes.html#cmdoption-arg-num_key_value_heads">command line option</a>
 </li>
       </ul></li>
       <li>