diff --git a/features/cli.html b/features/cli.html
index 8fe88a026..2cbdb6356 100644
--- a/features/cli.html
+++ b/features/cli.html
@@ -102,6 +102,7 @@
 <li class="toctree-l4"><a class="reference internal" href="#dataset-options,-which-mutually-exclusive-with-huggingface-dataset-options">dataset options, which mutually exclusive with huggingface dataset options</a></li>
 <li class="toctree-l4"><a class="reference internal" href="#huggingface-dataset-options,-if-dataset-options-are-not-provided,-user-should-provide-the-following-options-to-modify-the-default-data-config.-please-refer-to-olive.data.container.transformerstokendummydatacontainer-for-more-details.">huggingface dataset options, if dataset options are not provided, user should provide the following options to modify the default data config. Please refer to olive.data.container.TransformersTokenDummyDataContainer for more details.</a></li>
 <li class="toctree-l4"><a class="reference internal" href="#pass-options">pass options</a></li>
+<li class="toctree-l4"><a class="reference internal" href="#remote-options_repeat1">remote options</a></li>
 </ul>
 </li>
 <li class="toctree-l3"><a class="reference internal" href="#cloud-cache">cloud-cache</a><ul>
@@ -220,16 +221,16 @@ <h3>finetune<a class="headerlink" href="#finetune" title="Permalink to this head
 <p>Fine-tune a model on a dataset using peft and optimize the model for ONNX Runtime with adapters as inputs. Huggingface training arguments can be provided along with the defined options.</p>
 <div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">olive</span> <span class="n">finetune</span> <span class="p">[</span><span class="o">-</span><span class="n">h</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">precision</span> <span class="p">{</span><span class="n">float16</span><span class="p">,</span><span class="n">float32</span><span class="p">}]</span> <span class="o">-</span><span class="n">m</span> <span class="n">MODEL_NAME_OR_PATH</span>
                <span class="p">[</span><span class="o">--</span><span class="n">trust_remote_code</span><span class="p">]</span>
-               <span class="p">[</span><span class="o">--</span><span class="n">torch_dtype</span> <span class="p">{</span><span class="n">bfloat16</span><span class="p">,</span><span class="n">float16</span><span class="p">,</span><span class="n">float32</span><span class="p">}]</span> <span class="o">-</span><span class="n">d</span> <span class="n">DATA_NAME</span>
-               <span class="p">[</span><span class="o">--</span><span class="n">data_files</span> <span class="n">DATA_FILES</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">train_split</span> <span class="n">TRAIN_SPLIT</span><span class="p">]</span>
+               <span class="p">[</span><span class="o">--</span><span class="n">torch_dtype</span> <span class="p">{</span><span class="n">bfloat16</span><span class="p">,</span><span class="n">float16</span><span class="p">,</span><span class="n">float32</span><span class="p">}]</span> <span class="p">[</span><span class="o">--</span><span class="n">use_ort_genai</span><span class="p">]</span> <span class="o">-</span><span class="n">d</span>
+               <span class="n">DATA_NAME</span> <span class="p">[</span><span class="o">--</span><span class="n">data_files</span> <span class="n">DATA_FILES</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">train_split</span> <span class="n">TRAIN_SPLIT</span><span class="p">]</span>
                <span class="p">[</span><span class="o">--</span><span class="n">eval_split</span> <span class="n">EVAL_SPLIT</span><span class="p">]</span>
                <span class="p">(</span><span class="o">--</span><span class="n">text_field</span> <span class="n">TEXT_FIELD</span> <span class="o">|</span> <span class="o">--</span><span class="n">text_template</span> <span class="n">TEXT_TEMPLATE</span><span class="p">)</span>
                <span class="p">[</span><span class="o">--</span><span class="n">max_seq_len</span> <span class="n">MAX_SEQ_LEN</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">method</span> <span class="p">{</span><span class="n">lora</span><span class="p">,</span><span class="n">qlora</span><span class="p">}]</span>
                <span class="p">[</span><span class="o">--</span><span class="n">lora_r</span> <span class="n">LORA_R</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">lora_alpha</span> <span class="n">LORA_ALPHA</span><span class="p">]</span>
                <span class="p">[</span><span class="o">--</span><span class="n">target_modules</span> <span class="n">TARGET_MODULES</span><span class="p">]</span> <span class="p">[</span><span class="o">-</span><span class="n">o</span> <span class="n">OUTPUT_PATH</span><span class="p">]</span>
-               <span class="p">[</span><span class="o">--</span><span class="n">tempdir</span> <span class="n">TEMPDIR</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">clean</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">azureml_config</span> <span class="n">AZUREML_CONFIG</span><span class="p">]</span>
-               <span class="p">[</span><span class="o">--</span><span class="n">azureml_cluster</span> <span class="n">AZUREML_CLUSTER</span><span class="p">]</span>
-               <span class="p">[</span><span class="o">--</span><span class="n">azureml_keyvault</span> <span class="n">AZUREML_KEYVAULT</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">use_ort_genai</span><span class="p">]</span>
+               <span class="p">[</span><span class="o">--</span><span class="n">tempdir</span> <span class="n">TEMPDIR</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">clean</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">resource_group</span> <span class="n">RESOURCE_GROUP</span><span class="p">]</span>
+               <span class="p">[</span><span class="o">--</span><span class="n">workspace_name</span> <span class="n">WORKSPACE_NAME</span><span class="p">]</span>
+               <span class="p">[</span><span class="o">--</span><span class="n">keyvault_name</span> <span class="n">KEYVAULT_NAME</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">aml_compute</span> <span class="n">AML_COMPUTE</span><span class="p">]</span>
 </pre></div>
 </div>
 <section id="named-arguments_repeat1">
@@ -327,14 +328,17 @@ <h4>lora options<a class="headerlink" href="#lora-options" title="Permalink to t
 <section id="remote-options">
 <h4>remote options<a class="headerlink" href="#remote-options" title="Permalink to this heading">¶</a></h4>
 <dl class="option-list">
-<dt><kbd>--azureml_config</kbd></dt>
-<dd><p>Path to the azureml config file for remote run. Recommended to log in using <cite>az login</cite> before running.</p>
+<dt><kbd>--resource_group</kbd></dt>
+<dd><p>Resource group for the AzureML workspace.</p>
+</dd>
+<dt><kbd>--workspace_name</kbd></dt>
+<dd><p>Workspace name for the AzureML workspace.</p>
 </dd>
-<dt><kbd>--azureml_cluster</kbd></dt>
-<dd><p>The azureml cluster to use for remote run. Must be provided if using azureml_config.</p>
+<dt><kbd>--keyvault_name</kbd></dt>
+<dd><p>The azureml keyvault name with huggingface token to use for remote run. Refer to <a class="reference external" href="https://microsoft.github.io/Olive/features/huggingface_model_optimization.html#huggingface-login">https://microsoft.github.io/Olive/features/huggingface_model_optimization.html#huggingface-login</a> for more details.</p>
 </dd>
-<dt><kbd>--azureml_keyvault</kbd></dt>
-<dd><p>The azureml keyvault with huggingface token to use for remote run. Refer to <a class="reference external" href="https://microsoft.github.io/Olive/features/huggingface_model_optimization.html#huggingface-login">https://microsoft.github.io/Olive/features/huggingface_model_optimization.html#huggingface-login</a> for more details.</p>
+<dt><kbd>--aml_compute</kbd></dt>
+<dd><p>The compute name to run the workflow on.</p>
 </dd>
 </dl>
 </section>
@@ -502,6 +506,10 @@ <h3>tune-session-params<a class="headerlink" href="#tune-session-params" title="
                           <span class="p">[</span><span class="o">--</span><span class="n">disable_force_evaluate_other_eps</span><span class="p">]</span>
                           <span class="p">[</span><span class="o">--</span><span class="n">enable_profiling</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">output_path</span> <span class="n">OUTPUT_PATH</span><span class="p">]</span>
                           <span class="p">[</span><span class="o">--</span><span class="n">tempdir</span> <span class="n">TEMPDIR</span><span class="p">]</span>
+                          <span class="p">[</span><span class="o">--</span><span class="n">resource_group</span> <span class="n">RESOURCE_GROUP</span><span class="p">]</span>
+                          <span class="p">[</span><span class="o">--</span><span class="n">workspace_name</span> <span class="n">WORKSPACE_NAME</span><span class="p">]</span>
+                          <span class="p">[</span><span class="o">--</span><span class="n">keyvault_name</span> <span class="n">KEYVAULT_NAME</span><span class="p">]</span>
+                          <span class="p">[</span><span class="o">--</span><span class="n">aml_compute</span> <span class="n">AML_COMPUTE</span><span class="p">]</span>
 </pre></div>
 </div>
 <section id="named-arguments_repeat5">
@@ -631,6 +639,23 @@ <h4>pass options<a class="headerlink" href="#pass-options" title="Permalink to t
 </dd>
 </dl>
 </section>
+<section id="remote-options_repeat1">
+<h4>remote options<a class="headerlink" href="#remote-options_repeat1" title="Permalink to this heading">¶</a></h4>
+<dl class="option-list">
+<dt><kbd>--resource_group</kbd></dt>
+<dd><p>Resource group for the AzureML workspace.</p>
+</dd>
+<dt><kbd>--workspace_name</kbd></dt>
+<dd><p>Workspace name for the AzureML workspace.</p>
+</dd>
+<dt><kbd>--keyvault_name</kbd></dt>
+<dd><p>The azureml keyvault name with huggingface token to use for remote run. Refer to <a class="reference external" href="https://microsoft.github.io/Olive/features/huggingface_model_optimization.html#huggingface-login">https://microsoft.github.io/Olive/features/huggingface_model_optimization.html#huggingface-login</a> for more details.</p>
+</dd>
+<dt><kbd>--aml_compute</kbd></dt>
+<dd><p>The compute name to run the workflow on.</p>
+</dd>
+</dl>
+</section>
 </section>
 <section id="cloud-cache">
 <h3>cloud-cache<a class="headerlink" href="#cloud-cache" title="Permalink to this heading">¶</a></h3>
diff --git a/searchindex.js b/searchindex.js
index 85f720ad8..44cbcc789 100644
--- a/searchindex.js
+++ b/searchindex.js
@@ -1 +1 @@
-Search.setIndex({"docnames": ["api/engine", "api/evaluator", "api/metric", "api/models", "api/passes", "api/resource_path", "api/search-algorithms", "api/systems", "examples", "extending_olive/custom_scripts", "extending_olive/design", "extending_olive/how_to_add_optimization_pass", "features/azureml_integration", "features/cli", "features/cloud_model_cache", "features/conversion", "features/huggingface_model_optimization", "features/lora", "features/model_transformations_and_optimizations", "features/packaging_output_models", "features/passes/convert_onnx", "features/passes/convert_pytorch", "features/passes/onnx", "features/passes/openvino", "features/passes/pytorch", "features/passes/qnn", "features/passes/quant_onnx", "features/passes/quant_pytorch", "features/passes/snpe", "features/quantization", "features/run_workflow_remotely", "getstarted/installation", "getstarted/quickstart_bert_example", "getstarted/quickstart_examples", "getstarted/quickstart_resnet_example", "index", "overview/olive", "overview/options", "overview/quicktour", "tutorials/advanced_users", "tutorials/azure_arc", "tutorials/azureml_scripts", "tutorials/configure_auto_optimizer", "tutorials/configure_data", "tutorials/configure_metrics", "tutorials/configure_model_path", "tutorials/configure_pass", "tutorials/configure_systems"], "filenames": ["api/engine.rst", "api/evaluator.rst", "api/metric.rst", "api/models.rst", "api/passes.rst", "api/resource_path.rst", "api/search-algorithms.rst", "api/systems.rst", "examples.md", "extending_olive/custom_scripts.md", "extending_olive/design.md", "extending_olive/how_to_add_optimization_pass.md", "features/azureml_integration.md", "features/cli.rst", "features/cloud_model_cache.md", "features/conversion.rst", "features/huggingface_model_optimization.md", "features/lora.md", "features/model_transformations_and_optimizations.rst", "features/packaging_output_models.md", "features/passes/convert_onnx.md", "features/passes/convert_pytorch.md", "features/passes/onnx.md", "features/passes/openvino.md", "features/passes/pytorch.md", "features/passes/qnn.md", "features/passes/quant_onnx.md", "features/passes/quant_pytorch.md", "features/passes/snpe.md", "features/quantization.rst", "features/run_workflow_remotely.md", "getstarted/installation.md", "getstarted/quickstart_bert_example.md", "getstarted/quickstart_examples.rst", "getstarted/quickstart_resnet_example.md", "index.rst", "overview/olive.md", "overview/options.md", "overview/quicktour.md", "tutorials/advanced_users.md", "tutorials/azure_arc.md", "tutorials/azureml_scripts.md", "tutorials/configure_auto_optimizer.rst", "tutorials/configure_data.rst", "tutorials/configure_metrics.rst", "tutorials/configure_model_path.md", "tutorials/configure_pass.rst", "tutorials/configure_systems.rst"], "titles": ["Engine", "OliveEvaluator", "Metric", "OliveModels", "Passes", "ResourcePath", "SearchAlgorithms", "OliveSystems", "Examples", "Custom Scripts", "Design", "How to add new optimization Pass", "Azure ML integration", "Command Line Tools", "Cloud Model Cache", "Model Conversions", "Huggingface Integration", "LoRA Adapters", "Model Transformations and Optimizations", "Packaging Olive artifacts", "ONNX", "PyTorch", "ONNX", "OpenVINO", "PyTorch", "QNN", "ONNX", "PyTorch", "SNPE", "Model Quantizations", "Remote Workflow", "Installation", "BERT optimization with PTQ on CPU", "Quickstart Examples", "ResNet optimization with PTQ on CPU", "Welcome to Olive\u2019s documentation!", "Olive", "Olive Options", "Quick Tour", "Advanced User Tour", "Self-hosted Kubernetes cluster", "Azure ML scripts", "How To Configure Auto Optimizer", "How To Configure Data", "How To Configure Metric", "How To Set Model Path", "How To Configure Pass", "How To Configure System"], "terms": {"class": [0, 1, 2, 3, 4, 5, 9, 37, 43, 44, 46, 47], "oliv": [0, 1, 2, 3, 4, 5, 6, 7, 9, 10, 11, 14, 15, 16, 17, 18, 22, 24, 26, 27, 29, 31, 32, 34, 39, 41, 42, 43, 44, 45, 46, 47], "workflow_id": [0, 30, 37], "str": [0, 2, 3, 4, 5, 7, 9, 11, 14, 19, 37, 42, 43], "default_workflow": 0, "search_strategi": [0, 19, 37, 39, 42], "dict": [0, 2, 3, 4, 5, 7, 9, 11, 19, 37], "ani": [0, 2, 3, 4, 5, 7, 10, 11, 26, 37, 38, 43, 46], "searchstrategyconfig": 0, "none": [0, 2, 3, 4, 5, 7, 19, 37, 42, 43, 44, 47], "host": [0, 10, 11, 16, 19, 35, 42, 47], "systemconfig": [0, 7], "target": [0, 4, 8, 10, 13, 16, 19, 20, 25, 26, 36, 37, 38, 42, 44, 47], "evalu": [0, 1, 2, 3, 4, 9, 12, 13, 14, 16, 19, 42, 43, 44, 47], "oliveevaluatorconfig": [0, 39], "cache_dir": [0, 19, 37, 39, 42], "cach": [0, 4, 10, 19, 25, 30, 35, 37, 39, 42], "clean_cach": [0, 37], "bool": [0, 3, 4, 6, 7, 14, 19, 37, 42], "fals": [0, 3, 4, 6, 7, 11, 13, 14, 16, 19, 22, 37, 39, 42, 46, 47], "clean_evaluation_cach": [0, 37], "plot_pareto_fronti": [0, 37], "azureml_client_config": [0, 7, 47], "sourc": [0, 1, 2, 3, 4, 5, 7], "The": [0, 3, 4, 5, 6, 7, 10, 11, 13, 14, 16, 17, 18, 19, 20, 22, 23, 24, 25, 26, 28, 30, 36, 37, 38, 39, 41, 42, 43, 45, 47], "execut": [0, 3, 4, 11, 13, 14, 19, 22, 25, 26, 28, 37, 42, 47], "regist": [0, 10, 19, 37, 45], "step": [0, 4, 11, 14, 16, 22, 26, 30, 37, 38, 40], "It": [0, 4, 10, 11, 13, 16, 19, 22, 23, 24, 27, 30, 37, 38, 40, 46, 47], "facilit": 0, "output": [0, 4, 5, 9, 10, 11, 13, 14, 15, 17, 18, 19, 22, 24, 32, 34, 36, 37, 38, 39, 42, 43], "model": [0, 4, 5, 8, 10, 11, 17, 21, 24, 26, 27, 35, 36, 42, 43, 44, 47], "us": [0, 3, 4, 5, 6, 9, 10, 11, 13, 16, 17, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 31, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47], "provid": [0, 3, 4, 5, 9, 10, 11, 16, 17, 18, 19, 22, 24, 25, 26, 28, 30, 35, 36, 37, 38, 39, 41, 42, 43, 44, 46, 47], "criteria": 0, "produc": [0, 17, 18, 38], "": [0, 4, 9, 16, 17, 21, 36, 37, 38, 39, 43, 46], "pass_typ": 0, "type": [0, 2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 13, 16, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 37, 38, 39, 42, 43, 45, 46, 47], "pass": [0, 9, 12, 14, 16, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 32, 34, 35, 40, 42, 43, 47], "config": [0, 3, 4, 5, 10, 11, 12, 17, 20, 22, 23, 24, 25, 26, 27, 28, 30, 31, 38, 39, 40, 41, 42, 44, 45, 46, 47], "disable_search": [0, 11, 26, 37, 46], "name": [0, 2, 4, 5, 6, 7, 9, 11, 12, 14, 16, 19, 22, 23, 30, 37, 38, 39, 41, 42, 43, 44, 47], "olivesystem": [0, 35, 39], "evaluator_config": [0, 39], "clean_run_cach": [0, 37], "output_nam": [0, 3, 4, 5, 12, 16, 28, 37, 38, 39, 45], "configur": [0, 4, 5, 6, 9, 10, 12, 16, 17, 35, 37, 38, 39, 45], "so": [0, 4, 11, 14, 19, 20, 22, 25], "could": [0, 4, 11, 18, 37, 38, 47], "instanti": [0, 39], "later": [0, 25], "run": [0, 4, 10, 12, 14, 16, 17, 19, 20, 22, 23, 24, 25, 26, 28, 37, 38, 40, 42, 43, 44, 47], "input_model_config": 0, "modelconfig": [0, 3, 39], "accelerator_spec": [0, 11, 19], "list": [0, 2, 3, 4, 5, 6, 7, 10, 13, 16, 19, 22, 31, 37, 38, 39, 43, 46, 47], "acceleratorspec": [0, 11], "packaging_config": [0, 19, 37], "packagingconfig": [0, 19, 37], "output_dir": [0, 4, 19, 37, 42], "evaluate_input_model": [0, 37], "true": [0, 2, 4, 5, 6, 11, 13, 14, 16, 19, 20, 22, 26, 27, 28, 37, 42, 44, 46, 47], "log_to_fil": [0, 37], "log_severity_level": [0, 37, 38], "int": [0, 3, 4, 5, 6, 7, 11, 19, 37, 42], "1": [0, 4, 6, 7, 9, 12, 13, 16, 19, 20, 22, 26, 28, 37, 38, 39, 42, 43, 44, 45, 47], "cloud_cache_config": [0, 14], "cloudcacheconfig": 0, "all": [0, 3, 4, 8, 10, 13, 19, 20, 36, 37, 38, 39, 42, 43, 46, 47], "input": [0, 3, 4, 10, 11, 12, 13, 14, 15, 17, 18, 19, 22, 23, 24, 26, 28, 42, 43, 47], "one": [0, 4, 8, 10, 11, 15, 19, 22, 24, 28, 38, 39, 41, 44], "more": [0, 4, 5, 6, 10, 12, 14, 16, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 37, 38, 39, 41, 42, 43, 47], "candid": [0, 10, 19, 32, 34], "paramet": [0, 4, 5, 6, 10, 11, 13, 17, 20, 22, 23, 24, 25, 26, 27, 28, 37, 39, 43, 46], "acceler": [0, 4, 7, 11, 19, 20, 21, 22, 24, 25, 26, 28, 36, 37, 42], "spec": [0, 10, 47], "packag": [0, 10, 11, 13, 27, 31, 32, 34, 35, 37, 38, 47], "i": [0, 3, 4, 5, 6, 9, 10, 11, 13, 16, 17, 18, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 31, 32, 34, 35, 36, 37, 38, 40, 41, 42, 43, 44, 45, 46, 47], "zip": [0, 19, 28], "file": [0, 4, 5, 9, 11, 12, 13, 16, 17, 22, 24, 25, 26, 27, 28, 30, 37, 38, 39, 40, 41, 42, 47], "directori": [0, 4, 9, 10, 13, 16, 28, 32, 34, 37, 38, 43, 44, 45], "save": [0, 4, 13, 22, 32, 34, 37, 45], "prefix": [0, 19, 37], "log": [0, 4, 13, 14, 16, 30, 37], "sever": [0, 12, 37, 40, 43], "level": [0, 4, 13, 20, 30, 37, 38, 45], "logger": [0, 4], "cloud": [0, 35, 36], "return": [0, 4, 9, 10, 11, 37, 39, 44], "search": [0, 4, 6, 11, 13, 18, 37, 39, 42, 46], "strategi": [0, 4, 16, 37], "ar": [0, 3, 4, 5, 6, 7, 10, 11, 12, 14, 16, 18, 19, 20, 22, 23, 24, 26, 36, 37, 38, 39, 41, 47], "order": [0, 4, 19, 22, 36, 37, 39], "thei": [0, 13, 19, 37, 43, 47], "were": [0, 4, 19, 37], "final": [0, 4, 10, 11, 14, 37, 38, 39, 47], "_": [0, 19, 43], "_model": 0, "onnx": [0, 4, 8, 10, 13, 15, 17, 18, 19, 25, 27, 28, 29, 32, 34, 37, 38, 39, 42, 47], "json": [0, 4, 9, 12, 13, 17, 30, 34, 38, 39, 41, 42, 43, 44, 46, 47], "result": [0, 4, 20, 24, 37, 39, 44, 47], "_metric": 0, "footprint": [0, 23, 24, 26], "find": [0, 10, 12, 16, 20, 22, 24, 30, 37, 38], "configbas": [0, 2], "can": [0, 4, 9, 10, 11, 12, 13, 14, 16, 17, 18, 19, 20, 21, 22, 24, 25, 26, 27, 28, 30, 31, 32, 34, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47], "assign": [0, 4, 46], "dictionari": [0, 4, 10, 37, 38, 43, 46, 47], "kei": [0, 4, 9, 13, 16, 37, 43], "correspond": [0, 4, 13, 14, 22, 28, 32, 34], "field": [0, 2, 4, 5, 7, 11, 13, 42, 43, 44], "pydant": [0, 2, 4, 5, 7], "set": [0, 2, 4, 5, 7, 10, 11, 13, 14, 16, 19, 20, 22, 25, 26, 28, 35, 36, 37, 38, 39, 40, 42, 46], "path": [0, 3, 4, 5, 7, 9, 11, 12, 13, 16, 19, 25, 28, 30, 35, 37, 38, 40, 41, 43, 47], "execution_ord": [0, 10, 19, 37, 39, 42], "requir": [0, 2, 4, 5, 7, 10, 11, 13, 16, 19, 22, 24, 26, 27, 36, 37, 38, 39, 41, 42, 43, 45, 46, 47], "search_algorithm": [0, 10, 19, 37, 39, 42], "search_algorithm_config": [0, 19, 37, 42], "output_model_num": [0, 37], "stop_when_goals_met": [0, 37], "max_it": [0, 37], "max_tim": [0, 37], "system": [0, 11, 12, 14, 35, 38, 42], "system_config": [0, 7], "systemtyp": 0, "targetuserconfig": [0, 7], "create_system": [0, 7], "properti": [0, 7], "olive_managed_env": [0, 7, 47], "olive_evalu": [0, 1, 39], "type_arg": 0, "option": [0, 4, 5, 6, 10, 11, 12, 20, 22, 25, 35, 38, 39, 42, 43, 46, 47], "user_script": [0, 4, 22, 24, 26, 28, 37, 43, 44], "script_dir": [0, 3, 4, 37, 43], "metric": [0, 4, 9, 10, 35, 37, 38, 39, 42], "is_accuracy_drop_toler": 0, "create_evalu": 0, "olivemodelhandl": [0, 3, 4, 11, 39], "oliveevalu": [0, 35], "backend": [2, 4, 16, 25, 37, 42], "torch_metr": [2, 37], "sub_typ": [2, 9, 16, 37, 39, 42, 44], "submetr": 2, "user_config": [2, 37, 44], "data_config": [2, 4, 9, 16, 22, 23, 26, 27, 28, 37, 39, 42, 43, 44, 46], "dataconfig": [2, 4, 43, 44], "get_inference_set": 2, "framework": [2, 4, 23, 24, 25, 26, 28, 36, 39], "get_run_kwarg": 2, "get_sub_type_info": 2, "info_nam": 2, "no_priority_filt": 2, "callback": 2, "function": [2, 3, 4, 9, 35, 37, 39, 43, 44], "lambda": 2, "valu": [2, 4, 5, 7, 9, 10, 11, 13, 16, 19, 20, 22, 24, 26, 37, 39, 41, 42, 43, 44, 46], "an": [2, 3, 4, 5, 7, 8, 10, 11, 13, 14, 16, 17, 18, 19, 20, 21, 22, 24, 26, 27, 28, 36, 37, 38, 39, 41, 43, 44, 45, 47], "enumer": [2, 5, 7], "accuraci": [2, 4, 9, 10, 16, 19, 20, 26, 27, 36, 37, 38, 42], "custom": [2, 4, 8, 10, 13, 22, 24, 35, 37], "latenc": [2, 3, 4, 10, 19, 23, 36, 37, 38, 39, 42], "throughput": [2, 10, 36, 37], "accuracy_scor": [2, 9, 37, 44], "auroc": [2, 9, 37, 44], "f1_score": [2, 9, 37, 44], "perplex": [2, 4, 24], "precis": [2, 4, 13, 15, 21, 23, 26, 37, 42], "recal": [2, 37], "avg": [2, 19, 37, 39, 42, 44], "max": [2, 4, 5, 9, 13, 37, 42, 44], "min": [2, 4, 37, 42, 44], "p50": 2, "p75": 2, "p90": 2, "p95": 2, "p99": 2, "p999": 2, "float": [2, 4, 13, 23, 26, 37], "has_regression_go": 2, "follow": [3, 4, 5, 6, 7, 9, 10, 11, 12, 16, 19, 22, 28, 30, 31, 36, 37, 38, 40, 42, 43, 44, 47], "avail": [3, 4, 5, 6, 7, 11, 12, 13, 14, 16, 20, 25, 28, 31, 37, 39, 42, 43, 46, 47], "which": [3, 4, 10, 11, 16, 17, 19, 20, 22, 23, 24, 25, 26, 27, 37, 38, 39, 42, 43, 47], "creat": [3, 4, 5, 9, 10, 11, 13, 16, 19, 22, 27, 31, 37, 39, 41, 43, 47], "hfmodelhandl": [3, 4, 37], "model_path": [3, 12, 16, 19, 37, 38, 39, 42, 43, 45], "resourcepathconfig": 3, "resourcepath": [3, 35], "task": [3, 9, 13, 16, 22, 37, 38, 42, 43], "text": [3, 13, 16, 37, 42, 43], "gener": [3, 4, 6, 13, 16, 17, 18, 19, 20, 22, 28, 37, 38, 45, 47], "past": [3, 4, 13, 16, 37], "load_kwarg": [3, 37], "hfloadkwarg": 3, "io_config": [3, 16, 37, 38, 39], "ioconfig": 3, "adapter_path": [3, 13, 17], "model_attribut": [3, 19], "distributedhfmodelhandl": [3, 4], "model_name_pattern": 3, "num_rank": 3, "pytorchmodelhandl": [3, 4, 23, 37], "model_file_format": 3, "modelfileformat": 3, "pytorch_entire_model": 3, "model_load": 3, "callabl": [3, 4, 11], "model_script": 3, "dummy_inputs_func": 3, "besid": [3, 20, 43], "load": [3, 4, 9, 13, 22, 24, 25, 28, 37, 38, 39, 43, 44], "also": [3, 4, 10, 11, 13, 16, 18, 19, 20, 22, 24, 25, 26, 27, 28, 31, 37, 38, 43, 45, 46], "get": [3, 4, 5, 13, 16, 36, 47], "io": [3, 4, 6, 13, 16, 37, 40], "from": [3, 4, 6, 9, 10, 11, 13, 14, 15, 16, 17, 19, 20, 21, 22, 24, 25, 26, 27, 28, 32, 34, 37, 38, 39, 41, 43, 44, 45, 46, 47], "user": [3, 4, 10, 11, 17, 19, 20, 22, 26, 27, 35, 37, 38, 42, 43, 46, 47], "dummi": [3, 16, 43], "onnxmodelhandl": [3, 4, 11, 23, 37], "onnx_file_nam": 3, "inference_set": [3, 19, 37], "use_ort_extens": [3, 19], "external_initializers_file_nam": 3, "constant_inputs_file_nam": 3, "graph": [3, 4, 13, 20, 22, 25, 26, 37], "mixin": 3, "onnxepvalidatemixin": 3, "valid": [3, 4, 11, 22, 23, 37, 42, 43], "onnxgraphmixin": 3, "support": [3, 4, 9, 10, 11, 14, 15, 16, 21, 22, 23, 24, 26, 27, 29, 37, 38, 39, 42, 45, 46, 47], "oper": [3, 4, 5, 10, 13, 19, 20, 22, 23, 25, 26, 37], "distributedonnxmodelhandl": [3, 4], "openvinomodelhandl": [3, 4, 37], "main": [3, 4, 10, 31, 43], "respons": [3, 10, 13, 17], "snpemodelhandl": [3, 4, 37], "input_nam": [3, 4, 16, 20, 28, 37, 38, 39, 43], "input_shap": [3, 4, 16, 20, 28, 37, 38, 39, 43], "output_shap": [3, 28], "compositemodelhandl": [3, 4], "model_compon": 3, "model_component_nam": 3, "compositemodel": 3, "repres": [3, 11, 20, 22, 26, 27, 39, 47], "multipl": [3, 4, 8, 13, 18, 19, 20, 22, 29, 36, 44], "compon": [3, 4, 10, 22, 36], "onli": [3, 4, 8, 9, 10, 11, 14, 19, 21, 24, 26, 27, 28, 37, 41, 42, 43, 44, 47], "get_model_compon": 3, "iter": [3, 4, 10, 37], "child": 3, "whisper": [3, 4, 8, 22], "exampl": [3, 4, 14, 17, 32, 34, 36, 38, 39, 42, 43, 44, 46, 47], "ha": [3, 4, 10, 11, 14, 20, 22, 24, 31, 37, 39, 42, 45], "encod": [3, 4, 22], "decod": [3, 4, 22], "collect": [3, 9], "contain": [3, 4, 10, 14, 19, 23, 28, 37, 43, 46, 47], "should": [3, 4, 9, 11, 20, 37, 40, 41, 43], "have": [3, 4, 9, 10, 12, 14, 16, 19, 20, 22, 24, 37, 38, 44, 47], "same": [3, 4, 10, 11, 14, 16, 19, 20, 22, 26, 30, 37, 42, 47], "each": [4, 5, 6, 10, 11, 13, 19, 20, 22, 24, 37, 38, 39, 43, 47], "descript": [4, 5, 6, 11, 19, 20, 37, 38], "convert": [4, 13, 15, 18, 21, 22, 23, 25, 26, 27, 28, 37, 39, 43], "pytorch": [4, 8, 15, 18, 20, 25, 26, 29, 32, 37, 38, 39], "torch": [4, 13, 20, 21, 27, 38], "export": [4, 17, 19, 22, 30, 47], "cpu": [4, 7, 8, 11, 12, 13, 16, 19, 21, 24, 27, 31, 33, 37, 38, 47], "handler": 4, "hf": [4, 16, 17, 37, 43], "script": [4, 11, 19, 35, 37, 39, 40, 43], "other": [4, 5, 10, 11, 14, 16, 20, 22, 26, 38, 39, 40, 43, 46], "object": [4, 9, 10, 11, 22, 38], "import": [4, 9, 11, 16, 32, 34, 38, 39, 43, 44, 46, 47], "thi": [4, 5, 6, 9, 10, 11, 12, 13, 14, 16, 19, 20, 21, 22, 23, 24, 26, 27, 28, 30, 31, 32, 34, 35, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47], "pathlib": [4, 5], "default_valu": [4, 5, 6, 11, 26, 46], "searchable_valu": [4, 11, 26, 46], "depend": [4, 11, 13, 37, 39, 46, 47], "save_as_external_data": [4, 26], "serial": 4, "tensor": [4, 21, 22, 24, 27, 38], "data": [4, 9, 11, 19, 20, 26, 30, 35, 37, 44], "separ": [4, 9, 13, 17, 22, 31, 37], "instead": [4, 14, 20, 25, 28, 38, 47], "directli": [4, 9, 14, 37, 38, 43, 45], "larg": [4, 24, 37], "2gb": 4, "mai": [4, 14, 20, 22, 31, 38, 39, 40], "forc": [4, 13], "extern": [4, 22], "regardless": 4, "all_tensors_to_one_fil": [4, 26], "effect": 4, "If": [4, 5, 6, 11, 12, 13, 14, 16, 19, 20, 25, 26, 28, 30, 31, 37, 38, 41, 42, 43, 44, 45, 47], "specifi": [4, 9, 11, 12, 16, 19, 20, 24, 26, 37, 38, 40, 41, 42, 44, 45, 47], "external_data_nam": 4, "model_path_nam": 4, "size_threshold": 4, "threshold": [4, 37], "size": [4, 13, 14, 19, 20, 23, 24, 37, 41, 44], "when": [4, 6, 9, 10, 13, 14, 16, 19, 22, 23, 28, 37, 43, 46, 47], "To": [4, 10, 11, 12, 22, 30, 34, 35, 37, 38], "everi": [4, 36], "raw": [4, 43], "0": [4, 9, 12, 13, 16, 19, 22, 23, 24, 28, 37, 38, 39, 41, 42, 43, 44, 47], "1024": [4, 13, 22], "convert_attribut": 4, "non": [4, 22], "attribut": [4, 9, 16, 19, 47], "target_opset": [4, 20, 22, 37, 38, 39], "version": [4, 5, 7, 12, 13, 16, 19, 23, 31, 37, 45, 47], "default": [4, 5, 8, 10, 11, 14, 16, 19, 20, 24, 26, 30, 31, 37, 38, 39, 40, 41, 42, 43, 46, 47], "ai": [4, 8, 14, 17, 20, 23, 25, 28, 29, 30, 31, 37, 38, 47], "opset": [4, 20], "13": [4, 16, 20, 37, 38, 39], "use_dynamo_export": 4, "whether": [4, 11, 13, 14, 19, 22, 37, 43, 46], "dynamo_export": 4, "api": [4, 5, 13, 27, 36, 39], "past_key_value_nam": 4, "argument": [4, 37, 41, 43], "point": [4, 6, 10, 11, 22, 23, 25, 26, 28, 37], "For": [4, 11, 12, 13, 16, 17, 19, 20, 22, 37, 38, 39, 42, 43, 46, 47], "huggingfac": [4, 14, 24, 27, 35, 37, 38, 43], "past_key_valu": [4, 37], "basic": [4, 20, 22], "devic": [4, 10, 13, 22, 25, 37, 38, 42, 44, 47], "convers": [4, 8, 10, 18, 27, 32, 34, 35, 38, 39, 42], "e": [4, 13, 19, 22, 26, 31, 37, 42, 47], "g": [4, 13, 19, 22, 26, 42, 47], "cuda": [4, 8, 11, 13, 38, 42], "distributedhfmodel": 4, "torch_dtyp": [4, 13], "dtype": [4, 13, 17, 37], "cast": [4, 22], "befor": [4, 13, 14, 18, 26, 37, 40, 41, 43], "float32": [4, 13, 15, 17, 37], "float16": [4, 13, 15, 17, 37, 42], "parallel_job": 4, "number": [4, 5, 6, 13, 14, 19, 22, 27, 37, 41, 43], "parallel": [4, 13], "job": [4, 5, 16, 37, 40], "disabl": [4, 42], "merge_adapter_weight": 4, "merg": [4, 8, 22, 24, 37], "adapt": [4, 18, 24, 35, 37], "weight": [4, 8, 10, 13, 21, 22, 23, 24, 26, 27, 28, 37], "after": [4, 5, 13, 18, 22, 24, 32, 34, 37], "structur": [4, 6, 10, 21, 22, 24, 37], "consist": [4, 10], "base": [4, 11, 19, 22, 24, 27, 37, 42, 43, 47], "That": 4, "you": [4, 9, 12, 13, 14, 16, 19, 20, 22, 25, 27, 28, 30, 31, 32, 34, 35, 37, 38, 39, 40, 41, 43, 44, 45, 46, 47], "cannot": [4, 20, 22, 25, 28, 37, 38, 47], "some": [4, 5, 10, 11, 20, 22, 35, 37, 46, 47], "fine": [4, 8, 13, 17, 20, 22, 24], "tune": [4, 8, 10, 11, 17, 18, 24, 26, 27, 32, 36, 39], "save_metadata_for_token_gener": 4, "metadata": 4, "token": [4, 10, 13, 16, 24, 37, 43], "includ": [4, 11, 12, 13, 14, 19, 20, 22, 26, 30, 38, 43, 47], "generation_config": [4, 24], "relat": [4, 22, 37], "latest": [4, 22, 31], "21": 4, "optim": [4, 8, 10, 13, 16, 17, 20, 21, 23, 24, 27, 33, 35, 36, 37, 38, 39, 43, 47], "fuse": [4, 22, 26, 37], "node": [4, 13, 20, 22, 37, 41], "transform": [4, 8, 16, 17, 20, 21, 23, 24, 27, 32, 35, 37, 38, 43], "scenario": [4, 8, 19, 20, 22, 36, 37], "where": [4, 9, 12, 14, 17, 19, 20, 21, 22, 24, 37, 38, 41, 43, 47], "runtim": [4, 8, 10, 13, 19, 20, 22, 26, 28, 32, 37, 47], "doe": [4, 6, 22, 37, 42, 43, 47], "appli": [4, 16, 18, 19, 22, 24, 26, 37, 39], "time": [4, 14, 19, 20, 22, 23, 26, 30, 37, 41, 42, 47], "onnxruntim": [4, 11, 19, 20, 22, 29, 31, 37, 47], "model_typ": [4, 16, 22, 37], "bert": [4, 8, 16, 22, 33, 37, 42, 43], "gpt2": 4, "bert_tf": 4, "tf2onnx": 4, "bert_kera": 4, "keras2onnx": 4, "unet": 4, "vae": 4, "clip": 4, "stabl": [4, 6, 8, 31], "diffus": [4, 8], "num_head": [4, 37], "attent": [4, 37], "head": [4, 37], "num_key_value_head": 4, "hidden_s": [4, 37], "hidden": [4, 37], "optimization_opt": 4, "turn": [4, 10, 30], "off": [4, 24, 27, 30], "fusion": [4, 20], "fusion_opt": 4, "fusionopt": 4, "opt_level": [4, 42], "2": [4, 7, 13, 16, 17, 19, 21, 22, 24, 26, 27, 37, 38, 39, 41, 42, 43, 44, 47], "extend": 4, "99": [4, 19], "categor": [4, 11, 47], "use_gpu": 4, "flag": [4, 40, 41, 47], "gpu": [4, 7, 8, 11, 13, 20, 21, 22, 24, 27, 31, 42, 47], "infer": [4, 13, 20, 21, 22, 23, 24, 25, 26, 27, 28, 36, 37, 43, 44, 45, 47], "only_onnxruntim": 4, "python": [4, 5, 9, 10, 13, 19, 25, 28, 30, 31, 32, 34, 37, 38, 39, 41, 43, 44, 46], "might": [4, 10, 13, 26, 43], "caus": [4, 37], "failur": 4, "symbol": [4, 20], "shape": [4, 15, 22, 37, 38], "condit": [4, 11, 20], "parent": [4, 11, 37], "half": 4, "keep_io_typ": [4, 20], "keep": [4, 42], "origin": [4, 23, 24, 27, 42], "force_fp32_op": 4, "force_fp32_nod": 4, "force_fp16_input": 4, "even": [4, 27], "convert_float_to_float16": 4, "tool": [4, 22, 25, 26, 28, 35, 36, 37, 41, 42], "prefer": [4, 18, 20], "them": [4, 9, 16, 20, 22, 25, 26, 28, 41, 44], "use_gqa": 4, "replac": [4, 6, 10, 13, 37], "multiheadattent": 4, "groupqueryattent": 4, "input_int32": 4, "int32": [4, 22], "comput": [4, 8, 10, 16, 19, 22, 23, 26, 28, 37, 40, 41, 47], "select": [4, 10, 26, 34, 38, 40], "process": [4, 8, 9, 13, 14, 16, 18, 23, 24, 28, 30, 37, 38, 39, 43], "cpu_cor": [4, 13], "core": [4, 10, 13, 14, 25, 28, 36], "thread": [4, 13, 22], "io_bind": [4, 13, 19, 42], "enabl": [4, 13, 14, 16, 27, 31, 36, 37, 42], "iobind": [4, 13], "enable_cuda_graph": [4, 13, 42], "providers_list": [4, 13, 22], "cpuexecutionprovid": [4, 11, 19, 22, 47], "execution_mode_list": [4, 13], "between": [4, 5, 10, 12, 13, 16, 22, 27, 37, 42], "opt_level_list": [4, 13], "trt_fp16_enabl": [4, 13], "fp16": [4, 8, 13, 21, 27, 42], "mode": [4, 13, 22, 31, 37, 38], "tensorrt": [4, 11, 13, 21, 24], "intra_thread_num_list": [4, 13], "intra": [4, 13], "test": [4, 13, 28, 31, 37], "inter_thread_num_list": [4, 13], "inter": [4, 13], "extra_session_config": [4, 13, 19], "extra": [4, 13, 19, 31, 47], "session": [4, 22, 30], "dure": [4, 13, 22, 23, 24, 31, 37], "force_evaluate_other_ep": 4, "differ": [4, 9, 10, 13, 19, 22, 24, 25, 28, 37, 38, 42, 44, 47], "associ": [4, 13], "enable_profil": [4, 13, 22, 37], "profil": [4, 13], "min_positive_v": 4, "constant": [4, 22, 35], "against": [4, 47], "1e": 4, "07": 4, "max_finite_v": 4, "10000": 4, "left": 4, "use_symbolic_shape_inf": 4, "op_block_list": [4, 20], "op": [4, 18, 20, 37], "leav": [4, 20], "node_block_list": 4, "name_pattern": 4, "whose": 4, "match": 4, "pattern": [4, 20, 21, 24], "By": [4, 37], "defaultlook": 4, "logit": [4, 16, 20, 44], "mix": [4, 15, 37], "simplifiedlayernorm": [4, 20], "skipsimplifiedlayernorm": [4, 20], "relu": [4, 20], "add": [4, 14, 16, 19, 20, 30, 35, 43, 45], "atol": 4, "absolut": [4, 23], "toler": 4, "check": [4, 14, 22, 24, 26, 27, 28, 30, 46], "06": 4, "preprocess": [4, 26, 43], "quantiz": [4, 8, 10, 18, 27, 32, 34, 35, 37, 38, 39], "qnn": [4, 8, 13, 18, 20, 26, 37], "fuse_layernorm": 4, "reducemean": 4, "sequenc": [4, 13, 37], "singl": [4, 8, 13, 19, 22, 26, 27, 36, 43], "layernorm": [4, 20], "inputs_to_make_channel_last": 4, "transpos": 4, "channel": 4, "last": [4, 38, 40], "input0": 4, "n": [4, 13, 17], "c": [4, 13, 24, 25, 26, 37, 41], "d1": 4, "d2": 4, "dn": 4, "chang": [4, 22, 38], "updat": [4, 20, 35, 38, 39], "input0_chanfirst": 4, "potenti": [4, 14], "improv": [4, 18, 20, 22, 23, 24, 26, 27, 35, 37, 42, 44], "qdq": [4, 8, 26], "ep": [4, 8, 10, 26, 42, 47], "becaus": 4, "addit": [4, 17, 23, 26, 31, 38, 43, 45], "allow": [4, 12, 13, 16, 20, 22, 23, 30, 36, 43], "insert": [4, 18, 23, 26], "ort": [4, 17, 18, 26], "layout": [4, 22], "cancel": 4, "out": [4, 22, 24, 26, 27, 28, 40, 46, 47], "outputs_to_make_channel_last": 4, "output0": 4, "output0_chanfirst": 4, "overrid": [4, 16, 19, 37, 39], "pre": [4, 8, 18, 24, 26, 35, 37, 46, 47], "resolv": 4, "constraint": [4, 36], "being": [4, 47], "refer": [4, 5, 6, 12, 16, 20, 21, 22, 23, 24, 25, 26, 27, 28, 31, 36, 37, 38, 39, 43, 44, 47], "situat": 4, "certain": [4, 10, 16, 20, 46], "16": [4, 13, 17, 22, 24, 44], "bit": [4, 8, 23, 24, 25, 26, 27, 28], "standalon": 4, "rather": 4, "neighbor": 4, "well": [4, 17, 23, 36, 43], "specif": [4, 9, 10, 16, 18, 19, 20, 22, 25, 26, 36, 37, 39, 46], "problem": 4, "aris": 4, "here": [4, 9, 12, 14, 16, 20, 22, 37, 38, 40, 42, 43, 44, 47], "note": [4, 11, 12, 16, 19, 20, 23, 24, 25, 26, 28, 31, 37, 38, 39, 40, 43, 45, 47], "handl": [4, 39], "just": [4, 38, 43], "initi": [4, 5, 10, 13, 22, 24, 37, 39], "activ": [4, 10, 23, 24, 26, 27, 31, 42], "overrides_config": 4, "format": [4, 13, 18, 19, 20, 22, 25, 26, 37], "tensor_nam": 4, "quant_typ": [4, 24], "element_wise_binary_op": 4, "element": 4, "wise": 4, "binari": [4, 18, 37], "sub": [4, 10], "mul": 4, "div": 4, "dynam": [4, 8, 15, 26, 37, 38], "quant_mod": 4, "weight_typ": [4, 26, 37, 39, 46], "both": [4, 20, 26, 39, 42, 44, 45, 47], "static": [4, 8, 11, 26, 37], "qint8": [4, 26], "sign": 4, "8": [4, 7, 13, 16, 23, 24, 25, 26, 27, 28, 43, 47], "integ": [4, 22, 23], "quint8": [4, 26, 37, 39, 46], "unsign": 4, "op_types_to_quant": 4, "append_first_op_types_to_quantize_list": 4, "append": [4, 9, 18], "firstli": [4, 19], "appear": 4, "nodes_to_quant": 4, "nodes_to_exclud": 4, "exclud": [4, 22, 37], "per_channel": [4, 26, 46], "per": [4, 10], "tip": 4, "reduce_rang": [4, 46], "http": [4, 5, 6, 13, 14, 31, 37, 43], "doc": [4, 42], "perform": [4, 10, 13, 18, 19, 20, 24, 26, 27, 32, 37, 38], "html": [4, 6, 13], "reduc": [4, 22, 23, 24, 26, 27, 36], "rang": [4, 22], "7": 4, "vnni": 4, "machin": [4, 10, 16, 19, 20, 22, 26, 41, 45, 47], "especi": 4, "quant_preprocess": 4, "prepar": [4, 14, 16, 36], "sigmoid": 4, "nnapi": [4, 20], "activationsymmetr": 4, "symmetr": [4, 13, 17], "calibr": [4, 25, 26], "weightsymmetr": 4, "enablesubgraph": 4, "subgraph": 4, "current": [4, 10, 13, 14, 35, 37, 42, 43, 45, 47], "forcequantizenoinputcheck": 4, "latent": 4, "like": [4, 5, 14, 19, 21, 22, 24, 26, 37, 38, 41, 43], "maxpool": 4, "do": [4, 11, 16], "alreadi": [4, 17, 19], "alwai": [4, 37], "behavior": [4, 5], "matmulconstbonli": 4, "matmul": 4, "const": 4, "b": [4, 20, 22, 24, 26], "conditionaldefault": [4, 11], "olive_invalid_param_valu": 4, "extra_opt": 4, "pair": 4, "pleas": [4, 5, 11, 12, 14, 16, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 31, 32, 34, 37, 38, 40, 44, 47], "github": [4, 13, 31, 37, 43], "com": [4, 5, 12, 31, 37, 43, 47], "microsoft": [4, 5, 12, 13, 31, 37, 40, 43, 47], "blob": [4, 43], "py": [4, 9, 19, 22, 24, 26, 28, 34, 37, 43, 44], "detail": [4, 5, 6, 12, 16, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 36, 37, 38, 39, 40, 41, 47], "about": [4, 6, 11, 14, 16, 20, 22, 23, 24, 25, 26, 27, 28, 37, 38, 39], "overwritten": 4, "calibrate_method": [4, 26], "method": [4, 11, 13, 17, 23, 24, 26, 37], "minmax": 4, "entropi": 4, "calibrationmethod": 4, "percentil": 4, "avoid": [4, 38, 43], "quant_format": [4, 26], "qoper": 4, "quantizelinear": 4, "dequantizelinear": 4, "activation_typ": [4, 26], "specialparamvalu": 4, "invalid": [4, 11, 22], "prepare_qnn_config": 4, "suitabl": [4, 20, 36], "qnn_extra_opt": 4, "execution_provid": [4, 7, 11, 19, 37, 42, 44, 47], "get_qnn_qdq_config": 4, "init_overrid": 4, "copi": 4, "necessari": [4, 14, 16, 26, 30, 32, 34, 38, 47], "adjust": 4, "tensorquantoverrid": 4, "either": [4, 9, 26, 37, 47], "assum": [4, 16, 37], "axi": [4, 37], "must": [4, 10, 11, 13, 16, 22, 25, 26, 37, 43, 46, 47], "present": [4, 37], "first": [4, 10, 16, 19, 32, 34, 37, 39, 43], "quanttyp": 4, "scale": [4, 13, 26, 41], "zero_point": 4, "zero": [4, 26], "rmax": 4, "maximum": [4, 13, 17, 37, 41], "real": 4, "rmin": 4, "minimum": [4, 13, 37, 41], "A": [4, 10, 19, 25, 41, 47], "nest": 4, "anoth": [4, 11, 15, 18, 20, 37, 38, 42, 43], "recv_nod": 4, "consum": [4, 13], "add_qtype_convert": 4, "automat": [4, 9, 12, 13, 16, 22, 30, 37, 38, 42, 45], "entri": 4, "ensur": 4, "ex": 4, "up": [4, 24, 27, 36, 42], "surround": 4, "appropri": 4, "document": [4, 12, 16, 22, 28, 36, 37, 40, 44, 45, 46], "mixed_precision_overrides_util": 4, "further": 4, "techniqu": [4, 10, 11, 22, 24, 26, 27, 36, 37, 38, 39], "ignor": [4, 5, 11, 13, 19, 41, 43], "olive_ignored_param_valu": 4, "fix": [4, 10, 15, 25, 26, 28, 37, 46], "dim_param": 4, "dim_valu": [4, 20], "comma": [4, 13], "3": [4, 5, 13, 19, 20, 23, 24, 25, 26, 27, 28, 37, 38, 39, 43], "256": [4, 13, 17], "intel": [4, 8, 23, 29, 37, 42, 43], "neural": [4, 8, 23, 28, 29, 37], "compressor": [4, 8, 29, 37], "approach": [4, 23, 24, 26, 39], "onnxrt_trt_ep": 4, "onnxrt_cuda_ep": 4, "domain": 4, "auto": [4, 10, 11, 35, 39], "cv": 4, "object_detect": 4, "nlp": [4, 8], "recommendation_system": 4, "adaptor": 4, "explicitli": [4, 45], "detect": 4, "workspac": [4, 5, 10, 13, 16, 19, 37, 40, 41, 45, 47], "intermedi": [4, 10, 14], "histori": [4, 19], "store": [4, 14, 30, 37, 38, 45], "nc_workspac": 4, "datetim": 4, "now": [4, 23, 37, 38, 39, 42, 43], "strftime": 4, "y": 4, "m": [4, 10, 13, 17, 25, 28, 30, 32, 34, 38], "d_": 4, "h": [4, 13, 17], "recip": 4, "below": [4, 47], "smooth_quant": 4, "smooth": 4, "quant": 4, "smooth_quant_arg": 4, "fast_bias_correct": 4, "fast": 4, "bia": 4, "correct": 4, "weight_correct": 4, "gemm_to_matmul": 4, "gemm": 4, "graph_optimization_level": [4, 19], "disable_al": 4, "enable_bas": 4, "enable_extend": 4, "enable_al": 4, "first_conv_or_matmul_quant": 4, "conv": 4, "last_conv_or_matmul_quant": 4, "pre_post_process_quant": 4, "postprocess": 4, "add_qdq_pair_to_weight": 4, "optypes_to_exclude_output_qu": 4, "don": [4, 11, 12, 20, 22, 37], "t": [4, 11, 12, 19, 20, 22, 37, 45, 47], "optyp": 4, "dedicated_qdq_pair": 4, "dedic": 4, "quant_level": 4, "choos": [4, 22], "conserv": 4, "combin": [4, 10, 27, 42, 46], "master": [4, 24], "tuning_strategi": 4, "md": 4, "algorithm": [4, 6, 24, 26, 27, 29, 37], "excluded_precis": 4, "empti": [4, 37], "fp32": [4, 8, 42], "bf16": 4, "int8": [4, 8, 26, 42], "want": [4, 14, 16, 19, 20, 30, 31, 37, 38, 39, 43], "tuning_criterion": 4, "instanc": [4, 5, 13, 19, 22, 39, 41, 47], "tuningcriterion": 4, "In": [4, 10, 12, 17, 20, 22, 26, 37, 39, 40, 42, 43, 44, 47], "strategy_kwarg": 4, "timeout": [4, 5, 37], "max_trial": 4, "5": [4, 5, 11, 19, 24, 34, 37, 39], "awar": [4, 18, 27, 36, 37, 42], "weight_only_config": [4, 26], "inc": [4, 25], "op_type_dict": 4, "4": [4, 7, 8, 9, 11, 19, 21, 22, 24, 26, 27, 37, 43], "group_siz": 4, "scheme": [4, 24], "asym": 4, "rtn": 4, "calibration_sampling_s": [4, 26], "sampl": [4, 6, 10, 13, 37], "100": [4, 42], "weight_onli": [4, 26], "we": [4, 10, 22, 28, 31, 36, 37, 38, 39, 40, 42, 43, 46], "best": [4, 10, 19, 20, 22, 32, 34, 36, 37, 39, 42, 44, 46], "vai_q_onnx": 4, "viti": [4, 8, 29, 37], "input_nod": 4, "start": [4, 19, 36, 39, 43], "need": [4, 10, 12, 14, 16, 18, 22, 26, 28, 30, 31, 37, 39, 42, 43, 44], "output_nod": 4, "end": [4, 37], "optimize_model": 4, "deprec": 4, "soon": [4, 39], "NOT": [4, 42], "recommend": [4, 13, 28, 31], "make": [4, 14, 16, 20, 24, 26, 27, 36, 40, 42], "debug": 4, "loss": 4, "difficult": 4, "use_external_data_format": 4, "nonoverflow": [4, 26], "minms": 4, "need_layer_fus": 4, "layer": [4, 21, 24, 37], "enable_dpu": 4, "dpu": [4, 8], "addqdqpairtoweight": 4, "remain": 4, "post": [4, 8, 9, 18, 24, 37, 43], "command": [4, 14, 16, 17, 22, 30, 31, 32, 34, 35, 38, 39, 40, 41], "tool_command": [4, 22], "composit": 4, "invok": [4, 11, 13], "tool_command_arg": [4, 22], "prepostprocessor": [4, 22], "schema": [4, 37], "would": [4, 38, 42, 43], "imag": [4, 19, 22, 47], "data_typ": [4, 22], "uint8": [4, 22], "num_byt": [4, 22], "append_pre_post_processing_op": 4, "prepostprocessorinput": [4, 22], "beam": [4, 18], "whisperbeamsearch": 4, "contrib": 4, "17": 4, "els": 4, "beamsearch": 4, "no_repeat_ngram_s": [4, 22], "ngram": 4, "occur": 4, "onc": [4, 14, 23, 26], "use_vocab_mask": 4, "vocab_mask": 4, "use_prefix_vocab_mask": 4, "prefix_vocab_mask": 4, "use_forced_decoder_id": 4, "decoder_input_id": 4, "use_logits_processor": 4, "logits_processor": 4, "use_temperatur": 4, "temperatur": 4, "extract": [4, 18, 37], "make_input": [4, 22], "proto": 4, "exist": [4, 13, 14, 16, 19, 20, 22, 41, 47], "sessionopt": 4, "add_external_initi": [4, 22], "pack_input": [4, 22], "pack": [4, 13, 22, 27], "modul": [4, 8, 9, 13, 20, 21, 37, 43], "dynamic_lora_r": 4, "lora_r": [4, 13], "optional_input": 4, "dimens": [4, 20, 22, 37], "save_format": [4, 13], "common": [4, 7, 47], "util": [4, 36, 37], "weightsfileformat": 4, "numpi": [4, 13], "hug": [4, 21, 24], "face": [4, 21, 24], "target_modul": [4, 13], "r": [4, 13, 32, 34], "64": [4, 13, 17, 19], "32": [4, 13, 17, 23, 26, 38, 39], "lora_alpha": [4, 13, 24], "alpha": [4, 13], "lora_dropout": 4, "dropout": 4, "probabl": 4, "05": 4, "modules_to_sav": 4, "apart": 4, "trainabl": [4, 22, 24], "checkpoint": [4, 13], "train": [4, 8, 13, 16, 17, 18, 20, 22, 36, 37], "bfloat16": [4, 13, 24], "allow_tf32": 4, "tf32": 4, "amper": 4, "speed": [4, 14, 26, 27], "inform": [4, 6, 11, 12, 23, 24, 39, 42], "see": [4, 20, 22, 37, 38, 39, 40], "org": 4, "tensorfloat": 4, "train_data_config": [4, 24], "eval_data_config": 4, "training_arg": [4, 24], "double_qu": 4, "again": 4, "fp4": 4, "nf4": [4, 24], "compute_dtyp": [4, 24], "loftq_it": 4, "trainer": 4, "trainingargu": 4, "paged_adamw_32bit": 4, "learning_r": [4, 24], "0002": [4, 24], "learn": [4, 5, 14, 16, 19, 20, 21, 22, 23, 24, 26, 27, 28, 36, 41, 45, 47], "rate": 4, "adamw": 4, "gradient_checkpoint": 4, "gradient": [4, 24], "lr_scheduler_typ": 4, "schedul": 4, "better": [4, 37, 42], "than": [4, 26, 38, 39], "cosin": 4, "advantag": [4, 21, 39], "analysi": 4, "warmup_ratio": 4, "03": 4, "fraction": 4, "warmup": 4, "evaluation_strategi": 4, "eval_dataset": 4, "otherwis": [4, 13, 37, 47], "unless": 4, "epoch": 4, "report_to": 4, "integr": [4, 8, 20, 22, 26, 27, 35, 36], "report": 4, "dir": 4, "temp": 4, "overwrite_output_dir": 4, "overwrit": 4, "content": [4, 25], "continu": [4, 14, 30], "resume_from_checkpoint": 4, "folder": [4, 5, 11, 12, 13, 19, 37], "superced": 4, "found": [4, 14, 16, 20, 22, 24, 25, 28, 37, 38, 39, 41], "extra_arg": [4, 25], "keyword": [4, 37, 43], "create_training_arg": 4, "val_data_config": 4, "training_loop_func": [4, 24], "loop": [4, 8, 24], "ptl_modul": [4, 24], "lightningmodul": 4, "lightn": [4, 8, 24], "wai": [4, 15, 22, 24, 26, 43, 44], "encapsul": 4, "logic": [4, 9], "readthedoc": [4, 6], "en": [4, 5, 6], "lightning_modul": 4, "ptl_data_modul": [4, 24], "lightningdatamodul": 4, "datamodul": 4, "num_epoch": [4, 24, 34], "num_step": 4, "do_valid": 4, "over": [4, 6, 10, 37, 39], "modules_to_fus": 4, "qconfig_func": 4, "qconfig": [4, 24], "qat": [4, 8, 24], "ao": 4, "pytorch_lightn": 4, "seed": [4, 6, 19, 37, 42], "random": [4, 6, 10, 37], "checkpoint_path": 4, "tensorflow": [4, 13, 25, 26, 28, 37], "openvino": [4, 8, 18, 37, 38, 42, 47], "your": [4, 8, 9, 13, 14, 16, 19, 22, 25, 27, 28, 30, 31, 37, 38, 39, 40, 41, 43, 44, 45], "example_input_func": 4, "tf": 4, "ndarrai": 4, "compress_to_fp16": 4, "compress": [4, 26, 36], "extra_config": [4, 19], "2023": [4, 23], "openvino_docs_ov_converter_ug_conversion_opt": 4, "output_model": 4, "ov_model": 4, "special": [4, 17, 38], "preserv": 4, "distilbert": 4, "etc": [4, 11, 17, 22, 26, 37, 38, 42, 47], "modeltypeenum": 4, "preset": 4, "defin": [4, 9, 10, 12, 13, 19, 26, 43, 44, 46], "presetenum": 4, "ignored_scop": 4, "basic_quantization_flow": 4, "ignored_scope_typ": 4, "scope": 4, "ignorescopetypeenum": 4, "target_devic": 4, "cpu_spr": [4, 7], "vpu": [4, 7], "workflow": [4, 9, 10, 12, 13, 14, 19, 32, 34, 35, 38, 39, 45], "hardwar": [4, 7, 8, 10, 18, 20, 22, 23, 26, 27, 36, 37, 38, 47], "snpe": [4, 8, 13, 18, 20, 37, 38, 42], "dlc": [4, 28, 37], "sdk": [4, 18, 37], "tensorflowmodelhandl": [4, 23], "length": [4, 13, 20, 37], "input_typ": [4, 16, 37, 43], "member": [4, 11], "platform_sdk": 4, "qualcomm": [4, 8, 25, 28], "inputtyp": 4, "input_layout": 4, "inputlayout": 4, "develop": [4, 25, 26, 28, 31, 36, 42], "site": 4, "string": [4, 5, 11, 13, 16, 22, 37], "enable_cpu_fallback": 4, "priority_hint": 4, "low": [4, 24], "use_enhanced_quant": 4, "enhanc": 4, "featur": [4, 13, 14, 31, 38, 39], "determin": [4, 42, 44], "long": [4, 30, 37], "tail": 4, "distribut": [4, 23, 37], "enable_htp": [4, 28], "htp": 4, "window": [4, 14, 38], "htp_soc": 4, "soc": 4, "offlin": [4, 22], "tools_snp": 4, "bias_bitwidth": 4, "overwrite_cache_record": 4, "oliveol": 4, "snpedevic": 4, "12": [4, 7, 37], "input_list": [4, 25], "qnnmodelhandl": 4, "input_dim": [4, 20], "network": [4, 23, 24, 26, 28, 37], "224": [4, 20, 43], "quot": 4, "charact": 4, "space": [4, 6, 10, 11, 37, 38, 46], "line": [4, 16, 17, 22, 35, 37, 38, 39, 41], "data2": 4, "out_nod": 4, "out_1": 4, "out_2": 4, "show_unconsumed_nod": 4, "custom_io": 4, "bundl": 4, "publicresourc": 4, "topic": 4, "80": 4, "63442": 4, "50": [4, 17], "compil": [4, 21, 25, 36], "code": [4, 13, 17, 19, 24, 31, 32, 34, 38, 43], "librari": [4, 18, 20, 21, 22, 24, 26, 27, 37], "lib": 4, "lib_target": [4, 25], "build": [4, 10, 19, 31, 36, 47], "aarch64": 4, "android": 4, "x86_64": [4, 25], "linux": [4, 25, 38], "clang": [4, 25], "lib_nam": 4, "bin": [4, 47], "qnn_model": 4, "context": [4, 18, 24, 37, 47], "particular": 4, "binary_fil": 4, "extens": [4, 21, 22, 36, 37], "arxiv": 4, "ab": 4, "2301": 4, "00774": 4, "bloom": [4, 24], "gpt_neox": 4, "llama": [4, 8, 16, 17, 37, 43], "opt": [4, 24], "sparsiti": [4, 8, 21, 24], "two": [4, 10, 22, 26, 46, 47], "semi": [4, 24], "nvidia": 4, "blog": [4, 24], "blocksiz": 4, "mask": 4, "128": [4, 13, 17, 43], "percdamp": 4, "percentag": [4, 37], "averag": 4, "hessian": 4, "diagon": 4, "dampen": 4, "01": [4, 9, 23, 37, 42, 44], "min_lay": 4, "prune": [4, 24, 26], "id": [4, 5, 12, 13, 19, 38, 41, 45, 47], "max_lay": 4, "layer_name_filt": 4, "given": [4, 13, 36, 37, 42, 43], "affect": [4, 14], "expect": [4, 24, 42], "most": [4, 20, 22, 27, 36, 37, 42, 47], "pdf": 4, "2401": 4, "15024": 4, "calibration_data_config": [4, 24], "dataset": [4, 9, 17, 24, 25, 26, 38, 43], "calcul": [4, 26, 44], "calibration_nsampl": 4, "calibration_batch_s": 4, "batch": [4, 13, 19, 20, 37, 43, 44], "42": 4, "measur": [4, 37], "how": [4, 12, 16, 22, 35, 36, 37, 38, 39, 40], "much": 4, "slice": [4, 24], "round_interv": 4, "interv": [4, 5, 37], "round": 4, "final_orient": 4, "orient": 4, "choic": [4, 13], "pca": 4, "new": [4, 13, 16, 24, 35, 40, 41, 47], "rotat": [4, 27], "abl": [4, 37, 40], "llm": [4, 27], "2404": 4, "00456": 4, "input_model_dtyp": 4, "modeldtyp": 4, "hadamard": 4, "rotation_se": 4, "matrix": 4, "replic": 4, "paper": [4, 24, 27], "w_rtn": [4, 27], "w_gptq": 4, "gptq": [4, 26, 27, 37], "flash_attention_2": 4, "newer": 4, "gptq_damp": 4, "damp": 4, "factor": [4, 14, 42], "gptq_opt_scal": 4, "w_bit": [4, 27], "w_asym": 4, "asymmetr": [4, 13, 17], "w_groupsiz": 4, "group": [4, 5, 6, 13, 37, 41, 47], "groupwis": 4, "a_bit": [4, 27], "a_asym": 4, "a_clip_ratio": 4, "ratio": 4, "new_max": 4, "clip_ratio": 4, "a_groups": 4, "k_bit": [4, 27], "k_clip_ratio": 4, "k_groupsiz": 4, "v_bit": [4, 27], "v_clip_ratio": 4, "v_groupsiz": 4, "optimum": [4, 8, 16, 37], "kernel": 4, "nsampl": 4, "layers_block_nam": 4, "block": [4, 10, 13], "fill": [4, 40, 43], "link": [4, 8, 12, 24], "autogptq": [4, 29], "896d8204bc89a7cfbda42bf3314e13cf4ce20b02": 4, "auto_gptq": 4, "l19": 4, "l26": 4, "outside_layer_modul": 4, "nn": [4, 21], "inside_layer_modul": 4, "linear": [4, 21], "damp_perc": 4, "static_group": 4, "true_sequenti": 4, "sequenti": 4, "desc_act": 4, "sym": 4, "awq": [4, 27, 37], "autoawq": [4, 29], "q_group_siz": 4, "column": [4, 24], "big": [4, 27], "batch_siz": [4, 9, 13, 16, 20, 22, 37, 39, 42, 43, 44], "gemv": 4, "compat": [4, 22, 26], "exllama": 4, "duo_scal": 4, "w": 4, "x": 4, "modules_to_not_convert": 4, "llava": 4, "mixtral": 4, "gate": [4, 16], "export_compat": 4, "down": [4, 30, 41], "pack_model_for_onnx_convers": [4, 27], "spars": [4, 21, 24], "applic": [4, 21], "entir": [4, 20, 37], "instal": [4, 10, 11, 16, 19, 32, 34, 35, 38, 40, 47], "dataload": [4, 9, 43], "14": [4, 19, 20], "decoder_model": 4, "decoder_with_past_model": 4, "mean": [4, 16, 26, 37], "main_export": 4, "its": [4, 5, 6, 10, 16, 20, 22, 23, 24, 25, 26, 27, 28, 37, 44, 47], "via": [4, 12, 37], "strict": 4, "decoder_with_past": 4, "strictli": 4, "case": [4, 13, 20, 22, 25, 32, 34, 37, 42, 43, 46, 47], "ad": [4, 9, 16, 22, 37], "builder": 4, "genai": 4, "model_build": 4, "metadata_onli": 4, "int4_block_s": [4, 13, 17], "block_siz": 4, "int4": [4, 8, 20, 42], "accept": [4, 11, 38], "int4_accuracy_level": 4, "accuracylevel": 4, "exclude_emb": 4, "remov": [4, 13], "embed": 4, "exclude_lm_head": 4, "languag": [4, 17, 27, 37], "captur": 4, "place": [4, 9], "prerequisit": [4, 18], "correctli": 4, "resourc": [5, 7, 12, 13, 16, 19, 37, 41, 45, 47], "resource_path": 5, "azureml_datastor": [5, 12, 37, 45], "azureml_job_output": [5, 12, 45], "azureml_model": [5, 12, 45], "azuremlregistrymodel": 5, "azureml_registry_model": [5, 12, 16, 45], "string_nam": [5, 45], "local": [5, 7, 8, 10, 12, 13, 30, 37, 38, 39], "azureml": [5, 7, 8, 10, 13, 14, 19, 40, 41], "azureml_cli": [5, 12, 16, 45, 47], "client": [5, 47], "azuremlclientconfig": [5, 7], "azuremlcli": 5, "mlclient": [5, 45], "read_timeout": [5, 37], "max_operation_retri": [5, 37], "operation_retry_interv": [5, 37], "control": [5, 20], "creation": [5, 19, 37], "download": [5, 12, 14, 16, 24, 37, 40, 47], "subscription_id": [5, 12, 13, 16, 37, 41, 45, 47], "azur": [5, 13, 19, 35, 45, 47], "subscript": [5, 13, 37, 41, 47], "aml_config_path": [5, 13, 41], "resource_group": [5, 12, 13, 16, 37, 41, 45, 47], "workspace_nam": [5, 12, 13, 16, 37, 41, 45, 47], "60": [5, 24], "read": [5, 23, 37], "second": [5, 10, 13, 37, 41], "request": [5, 37], "retri": [5, 37], "doubl": [5, 37], "default_auth_param": [5, 37], "auth": [5, 37], "u": [5, 22, 35, 39], "ident": [5, 16, 37, 47], "defaultazurecredenti": [5, 37], "view": 5, "keyvault_nam": [5, 16, 37], "keyvault": [5, 13, 16, 37], "retriev": [5, 16, 37], "secret": [5, 16, 37], "get_workspace_config": 5, "create_cli": 5, "create_registry_cli": 5, "registry_nam": [5, 12, 13, 16, 45], "datastor": [5, 7, 30, 37, 47], "datastore_nam": [5, 12, 37, 45], "relative_path": [5, 12, 37, 45], "rel": [5, 12, 45], "datastore_url": 5, "url": [5, 14, 37], "job_nam": 5, "exhaust": [6, 10, 22, 37, 39], "grid": 6, "searcher": 6, "without": [6, 10, 14, 27, 30, 37], "num_sampl": [6, 19, 37, 42], "suggest": [6, 35], "rng": 6, "with_replac": 6, "tpe": [6, 10, 19, 37, 42], "tree": [6, 10, 37], "parzen": [6, 10, 37], "estim": [6, 10, 37], "optuna": 6, "tpesampl": 6, "underneath": 6, "sampler": 6, "multivari": 6, "decompos": 6, "docker": [7, 10, 19, 37, 40], "isolatedort": [7, 47], "pythonenviron": [7, 37, 47], "intel_myriad": 7, "npu": [7, 8, 25, 47], "hf_token": [7, 16, 37], "local_docker_config": [7, 47], "localdockerconfig": [7, 47], "is_dev": 7, "requirements_fil": [7, 19, 47], "image_nam": [7, 47], "dockerfil": [7, 47], "build_context_path": [7, 47], "build_arg": 7, "run_param": 7, "aml_comput": [7, 12, 37, 47], "aml_docker_config": [7, 12, 37, 47], "azuremldockerconfig": 7, "aml_environment_config": [7, 47], "azuremlenvironmentconfig": 7, "tag": [7, 47], "workspaceblobstor": [7, 30, 47], "instance_count": [7, 19, 47], "base_imag": [7, 12, 37, 47], "conda_file_path": [7, 12, 37, 47], "python_environment_path": [7, 47], "environment_vari": [7, 19, 47], "prepend_to_path": [7, 47], "kwarg": [7, 37, 43], "python_environ": [7, 47], "isolated_ort": [7, 47], "system_alia": 7, "azurend12ssystem": 7, "classvar": 7, "num_cpu": 7, "num_gpu": 7, "sku": [7, 19], "standard_nd12": 7, "system_typ": 7, "azurend24rssystem": 7, "24": 7, "standard_nd24r": 7, "azurend24ssystem": 7, "standard_nd24": 7, "azurend6ssystem": 7, "6": [7, 11, 13, 28], "standard_nd6": 7, "azurend96a100system": 7, "96": 7, "standard_nd96amsr_a100_v4": 7, "azurend96asystem": 7, "standard_nd96asr_v4": 7, "azurendv2system": [7, 37, 47], "40": 7, "standard_nd40rs_v2": 7, "surfaceprosystem1796": 7, "surface_pro_1796": 7, "surfaceprosystem1807": 7, "surface_pro_1807": 7, "llama2": 8, "qlora": [8, 13, 17, 18, 22], "mistral": 8, "open": [8, 20, 22, 23, 26], "sparsegpt": [8, 18], "torchtrt": 8, "phi2": 8, "slicegpt": [8, 18], "falcon": 8, "red": 8, "pajama": 8, "trt": [8, 42], "deberta": 8, "registri": [8, 9, 13, 43], "gptj": 8, "audio": 8, "spectrogram": 8, "vision": 8, "xl": 8, "directml": [8, 11, 31, 47], "toolkit": [8, 23], "squeezenet": 8, "mobilenet": 8, "resnet": [8, 33, 37, 38, 39], "amd": [8, 29, 37], "vgg": 8, "incept": [8, 43], "super": 8, "resolut": 8, "abil": [9, 26], "write": [9, 22, 23, 39, 42], "own": [9, 10, 13, 16, 22, 30, 36, 37, 43, 44, 45], "evaluate_func": [9, 37, 44], "metric_func": [9, 37, 44], "openvinoquant": [9, 23, 37], "my_script": 9, "post_process": [9, 37, 43, 44], "mydataload": 9, "def": [9, 11, 43, 44], "__init__": 9, "self": [9, 11, 12, 35], "__len__": 9, "__getitem__": 9, "register_dataload": [9, 43], "my_dataload": 9, "register_post_process": [9, 43], "my_post_process": 9, "accuracy_data_config": [9, 37, 44], "load_dataset_config": [9, 16, 42, 43], "skip_dataset": 9, "post_process_data_config": [9, 37, 43, 44], "dataloader_config": [9, 22, 24, 26, 28, 37, 42, 43, 44], "param": [9, 11, 22, 37, 43, 44], "prioriti": [9, 16, 19, 37, 39, 42, 44], "goal": [9, 10, 26, 37, 42, 44], "degrad": [9, 37, 42, 44], "sy": 9, "interpret": 9, "let": [9, 38, 39, 43, 46], "sai": [9, 10, 11], "my_modul": 9, "my_olive_project": 9, "print": 9, "my_val": 9, "section": [10, 12, 16, 37, 43, 45], "discuss": 10, "concept": [10, 47], "compos": [10, 36, 37, 43], "modular": 10, "construct": 10, "achiev": [10, 26], "respect": [10, 26], "emploi": 10, "togeth": [10, 22, 37], "desir": [10, 25, 26, 31, 36, 38, 39], "pareto": [10, 37], "frontier": [10, 37], "multi": 10, "meet": [10, 43], "diagram": 10, "illustr": 10, "relationship": 10, "searchabl": [10, 11, 26, 37, 46], "chose": [10, 39, 46], "possibl": [10, 13, 20, 27, 38, 46, 47], "environ": [10, 13, 19, 20, 22, 25, 26, 28, 30, 31, 37, 38, 40], "o": [10, 13, 17, 38, 47], "platform": [10, 20, 22, 23, 26, 47], "thu": [10, 22, 26, 37], "ideal": 10, "deploy": [10, 19, 36], "five": [10, 47], "localsystem": [10, 37, 39, 42, 47], "azuremlsystem": 10, "dockersystem": [10, 47], "pythonenvironmentsystem": [10, 47], "isolatedortsystem": [10, 47], "light": 10, "three": [10, 38, 46], "tp": 10, "along": [10, 11, 13, 24, 38, 43], "manag": [10, 12, 16, 37, 41], "maintain": 10, "pipelin": [10, 30, 32, 37], "joint": [10, 19, 37, 39, 42], "independ": 10, "trial": 10, "been": [10, 14, 22, 39], "implement": [10, 13, 22, 24, 26, 27, 28, 36, 37, 38], "randomli": 10, "ampl": 10, "simpl": [11, 19, 38, 39, 42], "interfac": [11, 17, 20, 22, 26, 39, 41], "introduc": [11, 22, 35], "newoptimizationtrick": 11, "next": [11, 14, 19], "_default_config": 11, "take": [11, 21, 30, 36, 37, 38, 39, 40, 43, 44, 46], "passconfigparam": 11, "dataclass": 11, "hold": 11, "accelerator_typ": 11, "cudaexecutionprovid": [11, 22, 42, 47], "belong": 11, "dml": 11, "sinc": [11, 14, 19, 22, 36, 46, 47], "type_": 11, "categori": 11, "upload": [11, 12, 14, 19], "normal": [11, 30, 47], "root": [11, 12, 13, 45], "searchparamet": 11, "classmethod": 11, "cl": 11, "param1": 11, "param2": 11, "param3": 11, "objectclass": 11, "param4": 11, "union": [11, 37], "paramcategori": 11, "param5": 11, "param6": 11, "_run_for_config": 11, "engin": [11, 12, 19, 21, 25, 28, 35, 36, 40, 42, 47], "while": [11, 20, 22, 26, 36], "receiv": [11, 18, 37], "output_model_path": 11, "outlin": [12, 16], "discov": [12, 16], "asset": [12, 19, 30, 37, 41], "within": [12, 16], "aml": [12, 16, 19, 37, 41], "model_nam": [12, 13, 19, 43, 45], "model_vers": [12, 19], "my_subscription_id": [12, 45], "my_resource_group": [12, 45], "my_workspac": [12, 45], "my_datastor": [12, 45], "model_dir": [12, 45], "my_model": [12, 45], "pt": [12, 13, 38, 39, 45], "job_id": [12, 45], "my_job_id": [12, 45], "my_output_nam": [12, 45], "data_dir": [12, 37, 43, 44], "aml_system": [12, 16, 30, 37], "cluster": [12, 13, 30, 35, 37, 47], "mcr": [12, 37, 47], "openmpi4": [12, 37, 47], "ubuntu20": [12, 37, 47], "04": [12, 37, 47], "conda": [12, 31, 37, 47], "yaml": [12, 13, 17, 37, 47], "Then": [12, 16, 30, 32, 34, 40, 43], "azuremlmodel": 12, "azuremldata": 12, "instruct": [12, 16, 28, 30, 36, 40], "setup": [12, 13, 32, 34, 38], "kubernet": [12, 35], "offer": [12, 36], "assist": [12, 37], "variou": [13, 15, 22, 36, 38], "usag": [13, 16, 24, 35], "package_config": 13, "run_config": 13, "tempdir": [13, 38], "advanc": [13, 35], "locat": [13, 14, 37, 41], "individu": [13, 39], "proprietari": [13, 38], "privat": [13, 38], "tempfil": 13, "peft": [13, 17, 22, 24], "model_name_or_path": 13, "trust_remote_cod": [13, 43], "d": [13, 17, 26, 41], "data_nam": [13, 16, 42, 43], "data_fil": [13, 43], "train_split": [13, 17], "eval_split": [13, 17], "text_field": 13, "text_templ": [13, 17], "max_seq_len": [13, 16], "output_path": [13, 17], "clean": [13, 37], "azureml_config": 13, "azureml_clust": 13, "azureml_keyvault": 13, "use_ort_genai": 13, "trust": [13, 43], "onnxrunti": 13, "split": [13, 16, 42, 43], "templat": [13, 37, 42], "question": [13, 17], "prompt": [13, 17, 43], "answer": [13, 17], "az": [13, 14], "login": [13, 14], "huggingface_model_optim": 13, "extractedadapt": 13, "safetensor": 13, "pack_weight": [13, 17], "quantize_int4": [13, 17], "int4_quantization_mod": [13, 17], "Will": 13, "arrai": 13, "blockwis": 13, "py_vers": [13, 25, 28], "15": [13, 28], "delet": [13, 41], "compute_nam": [13, 41], "vm_size": [13, 41], "min_nod": [13, 41], "max_nod": [13, 41], "idle_time_before_scale_down": [13, 41], "vm": [13, 41], "idl": [13, 41], "scaledown": 13, "120": [13, 41], "simpli": [13, 32, 34, 43], "onnx_model_path": 13, "hf_model_nam": 13, "device_typ": 13, "data_config_path": 13, "predict_with_kv_cach": 13, "seq_len": [13, 43], "past_seq_len": [13, 43], "shared_kv": [13, 37], "ort_past_key_nam": [13, 37], "ort_past_value_nam": [13, 37], "max_sampl": [13, 42, 43], "fields_no_batch": 13, "disable_force_evaluate_other_ep": 13, "perf_tuning_output": 13, "perf_tun": [13, 16], "share": [13, 14, 37], "kv": 13, "sensit": 13, "key1": 13, "value1": 13, "key2": 13, "value2": 13, "account": [13, 14, 16, 37], "model_hash": 13, "hash": 13, "whenev": 14, "With": [14, 16, 27, 31, 36, 42], "easili": [14, 36, 37, 43], "nativ": [14, 18], "pip": [14, 23, 30, 38, 47], "ll": 14, "readi": [14, 18, 36, 39], "page": [14, 16, 40], "typic": [14, 18, 38, 47], "look": [14, 38, 39, 42, 43], "account_url": 14, "net": 14, "sure": [14, 16, 40], "enable_cloud_cach": 14, "container_nam": 14, "upload_to_cloud": 14, "olivecachemodel": 14, "immedi": 14, "skip": [14, 20, 37], "proce": 14, "until": 14, "complet": [14, 16, 24, 37, 38, 39, 43], "faster": [14, 25, 26, 27], "slower": 14, "influenc": 14, "internet": 14, "mani": [14, 20], "form": [15, 22, 25, 37], "torchtrtconvers": 15, "hfmodel": [16, 21, 24, 37, 42], "input_model": [16, 37, 38, 39, 42, 45], "meta": [16, 17, 37, 43], "7b": [16, 17, 37, 43], "catalog": 16, "futur": 16, "simplifi": [16, 17, 36, 47], "fetch": 16, "onnxconvers": [16, 20, 37, 38, 39, 42], "onnxconfig": 16, "manual": [16, 42], "input_id": [16, 37, 43], "attention_mask": [16, 43], "position_id": 16, "int64": [16, 22, 43], "dynamic_ax": [16, 37, 39], "sequence_length": [16, 37], "total_sequence_length": 16, "huggingfacecontain": [16, 42, 43], "oasst1_train": 16, "timdettm": 16, "openassist": 16, "guanaco": 16, "pre_process_data_config": [16, 42, 43], "text_col": 16, "512": 16, "pad_to_max_len": 16, "ortperftun": [16, 22, 37, 42], "huggingface_metr": [16, 37, 42], "f1": [16, 37, 42], "access": [16, 37, 40], "cli": 16, "termin": [16, 30], "vault": 16, "my_keyvault_nam": 16, "reserv": 16, "purpos": 16, "servic": 16, "msi": 16, "grant": 16, "guid": [16, 38, 39], "abov": [16, 17, 20, 24, 32, 34, 37, 42, 43, 47], "ptq": [16, 18, 33], "standard": [17, 39], "finetun": [17, 24], "usual": [17, 26], "power": [17, 26, 30, 42], "ux": 17, "v": 17, "flexibl": [17, 20, 22, 26, 30, 42], "describ": [17, 22, 24, 38, 44, 45, 46], "llama2_qlora": 17, "altern": [17, 19, 20, 44], "nampdn": 17, "tini": 17, "4096": 17, "4224": 17, "per_device_train_batch_s": 17, "per_device_eval_batch_s": 17, "max_step": 17, "150": 17, "logging_step": 17, "programming_languag": 17, "consumpt": 17, "known": [18, 19, 28, 38], "goe": 18, "through": [18, 22, 24, 26, 37, 42], "seri": 18, "product": 18, "design": [18, 35], "lora": [18, 22, 35], "loftq": 18, "mergeadapterweight": [18, 37], "There": [19, 26, 35, 37, 43, 47], "samplecod": 19, "onnxruntimepackag": 19, "models_rank": 19, "top": [19, 30, 37, 45], "across": [19, 22, 26, 36], "k": 19, "bestcandidatemodel_1": 19, "bestcandidatemodel_2": 19, "bestcandidatemodel_k": 19, "metric_1": 19, "metric_2": 19, "metric_3": 19, "sort": 19, "lower": [19, 24], "bestcandidatemodel": 19, "info": [19, 37], "model_config": 19, "onnxmodel": 19, "provider_opt": 19, "session_opt": [19, 37], "execution_mod": 19, "inter_op_num_thread": 19, "intra_op_num_thread": 19, "model_attributes_kei": 19, "model_attributes_valu": 19, "8602941176470589": 19, "higher_is_bett": [19, 37, 44], "36": 19, "2313": 19, "packaging_config_nam": 19, "model_rank": 19, "olive_output_model": 19, "olive_output_model_cpu": 19, "cpu_1": 19, "ml": [19, 25, 30, 35, 36, 45, 47], "No": 19, "endpoint": 19, "doesn": [19, 45, 47], "deploi": [19, 23, 40], "packagingtyp": 19, "outputmodel": 19, "export_in_mlflow_format": 19, "mlflow": 19, "registr": 19, "model_packag": 19, "modelpackageconfig": 19, "target_environ": 19, "target_environment_vers": 19, "plu": 19, "inferencing_serv": 19, "inferenceserverconfig": 19, "inferenc": [19, 22], "server": 19, "azuremlonlin": 19, "azuremlbatch": 19, "code_fold": 19, "score": 19, "scoring_script": 19, "base_environment_id": 19, "env": 19, "var": 19, "necessarili": 19, "deployment_config": 19, "deploymentconfig": 19, "endpoint_nam": 19, "deployment_nam": 19, "instance_typ": 19, "managedonlinedeploy": 19, "batchdeploy": 19, "interf": 19, "mini_batch_s": 19, "mini": 19, "invoc": 19, "10": [19, 37], "txt": [19, 25, 32, 34, 38, 43, 47], "include_runtime_packag": 19, "common_evalu": [19, 37, 42, 43], "local_system": [19, 37, 39, 42, 47], "null": [19, 37, 43], "53fc6781998a4624b61959bb064622c": 19, "0_onnxconvers": 19, "7a320d6d630bced3548f242238392730": 19, "1_orttransformersoptim": 19, "c499e39e42693aaab050820afd31e0c3": 19, "2_onnxquant": 19, "1431c563dcfda9c9c3bf26c5d61ef58": 19, "3_ortperftun": 19, "a843d77ae4964c04e145b83567fb5b05": 19, "cross": [20, 22, 23, 26], "onnxopversionconvers": [20, 37], "similar": 20, "modelbuild": [20, 37], "decreas": 20, "onnxfloattofloat16": 20, "often": 20, "expos": 20, "stage": 20, "stand": [20, 23], "alon": 20, "suit": 20, "architectur": [20, 25], "reli": 20, "grain": 20, "webgpu": 20, "onnxiofloat16tofloat32": 20, "give": [20, 35, 37], "poor": 20, "ortmixedprecis": 20, "minim": 20, "retain": 20, "skiplayernorm": 20, "fastgelu": 20, "embedlayernorm": 20, "mobil": 20, "dynamictofixedshap": [20, 37], "effici": [20, 23, 24, 26, 36], "coreml": 20, "helper": 20, "exclus": [20, 41], "deep": [21, 23, 24, 26, 27, 28, 43], "trtmodul": 21, "torch_tensorrt": 21, "regular": 21, "onnxmodeloptim": [22, 37], "involv": 22, "cost": [22, 26], "analyz": 22, "identifi": [22, 38], "argmax": 22, "isn": 22, "tensorproto": 22, "yet": 22, "orttransformersoptim": [22, 37, 42], "capabl": 22, "These": [22, 26, 39, 47], "done": 22, "appendprepostprocessingop": 22, "superresolut": 22, "output_format": 22, "png": 22, "use_audio_decod": 22, "leverag": [22, 43], "placehold": 22, "__model_input__": 22, "__model_output__": 22, "iomapentri": 22, "full": [22, 38, 46, 47], "ycbcrtopixel": 22, "bgr": 22, "io_map": 22, "y1_uint8": 22, "cb1_uint8": 22, "cr1_uint8": 22, "knowledg": [22, 42], "built": [22, 37, 42], "convertimagetobgr": 22, "incompat": 22, "web": 22, "kind": [22, 40, 47], "proper": [22, 24], "exactli": 22, "resiz": 22, "resize_to": 22, "input_index": 22, "dim_index": 22, "centercrop": 22, "height": 22, "width": [22, 24], "pixelstoycbcr": 22, "imagebytestofloat": 22, "unsqueez": 22, "ax": [22, 37], "squeez": 22, "floattoimagebyt": 22, "output_index": 22, "hw": 22, "multipli": 22, "convertbgrtoimag": 22, "image_format": 22, "insertbeamsearch": 22, "chain": [22, 36], "high": 22, "cover": [22, 43], "knob": 22, "perf_tuning_data_config": 22, "device_id": 22, "arena_extend_strategi": 22, "knextpoweroftwo": 22, "gpu_mem_limit": 22, "2147483648": 22, "cudnn_conv_algo_search": 22, "do_copy_in_default_stream": 22, "calib_data_config": [22, 23, 26, 39, 46], "small": [22, 37], "matric": [22, 24], "call": [22, 24, 26, 27, 39, 42, 43, 44], "extractadapt": [22, 37], "modifi": 22, "itself": 22, "add_initi": 22, "part": [22, 38], "As": [22, 26, 47], "visual": 23, "focus": [23, 26], "anywher": 23, "openvinoconvers": [23, 37], "ir": [23, 37, 42], "uniform": 23, "move": 23, "help": [23, 41, 42], "memori": [23, 24, 26, 27], "arithmet": 23, "undergo": 23, "actual": 23, "transit": 23, "happen": 23, "openvinoquantizationwithaccuraci": 23, "validation_func": 23, "max_drop": 23, "drop_typ": 23, "rank": [24, 44], "freez": 24, "inject": 24, "decomposit": 24, "hftrainingargu": 24, "backpropag": 24, "frozen": 24, "pretrain": 24, "simultan": 24, "fake": 24, "quantizationawaretrain": [24, 37], "ptldatamodul": 24, "ptlmodul": 24, "gpt": 24, "shot": 24, "unstructur": 24, "175b": 24, "176b": 24, "neglig": 24, "increas": [24, 26, 37], "sparsif": 24, "smaller": 24, "orthogon": 24, "least": [24, 43], "signific": 24, "row": 24, "speedup": 24, "wikitext2": 24, "direct": 25, "technologi": 25, "softwar": [25, 28], "chipset": 25, "variabl": [25, 28, 37, 40, 47], "qnn_sdk_root": 25, "qnnconvers": 25, "qnnmodellibgener": 25, "qnncontextbinarygener": 25, "qnn_backend": 25, "higher": [26, 37], "compar": [26, 27], "ones": 26, "consolid": [26, 27], "onnxquant": [26, 37, 39, 46], "hyperparamet": [26, 27], "onnxdynamicquant": [26, 37, 38], "onnxstaticquant": [26, 37], "qnnpreprocess": 26, "unsupport": 26, "olivepass": 26, "onnx_quant": [26, 37, 46], "popular": [26, 43], "mxnet": 26, "varieti": [26, 39], "distil": 26, "experi": [26, 36], "driven": 26, "friendli": [26, 27], "easi": [26, 27, 36, 42], "strive": 26, "reach": 26, "everywher": 26, "incquant": [26, 37], "incdynamicquant": [26, 37], "incstaticquant": [26, 37], "inc_quant": 26, "complex": 26, "less": 26, "bandwidth": 26, "vitisaiquant": [26, 37], "vitis_ai_quant": 26, "favorit": 27, "come": 27, "drop": 27, "gptqquantiz": [27, 37], "trade": 27, "wikitext2_train": 27, "3x": 27, "wa": 27, "upon": 27, "work": [27, 36, 37], "mit": 27, "autoawqquant": [27, 37], "snapdragon": 28, "hexagon": 28, "dsp": 28, "v2": 28, "18": 28, "240101": 28, "snpe_root": 28, "snpeconvers": [28, 37], "299": 28, "inceptionv3": 28, "predict": [28, 37, 43], "reshape_1": 28, "1001": 28, "snpequant": [28, 37], "represent": 28, "32bit": 28, "bias": 28, "quant_data_config": 28, "quarot": 29, "awai": 30, "enough": 30, "shut": 30, "interrupt": 30, "workflow_host": [30, 37], "configu": 30, "config_fil": [30, 42], "submit": 30, "safe": 30, "close": 30, "artifact": [30, 35, 37, 47], "yyyi": 30, "mm": 30, "dd": 30, "hh": 30, "ss": 30, "date": 30, "virtual": [31, 47], "public": 31, "pypi": 31, "git": 31, "egg": 31, "contribut": 31, "clone": [31, 38], "repositori": [31, 32, 34], "cd": 31, "tuner": [32, 34], "go": [32, 34], "quickstart": [32, 34, 35], "accord": [32, 34], "olive_run": [32, 34, 38, 39], "bert_auto": 32, "prepare_model_data": 34, "resnet_ptq_cpu": 34, "under": [35, 37, 42, 43], "releas": 35, "feel": 35, "free": [35, 40], "feedback": 35, "quick": [35, 39, 42], "tour": 35, "remot": [35, 38, 43, 47], "olivemodel": [35, 39], "searchalgorithm": 35, "industri": 36, "lead": [36, 39], "edg": 36, "consider": 36, "vendor": 36, "fragment": 36, "effort": 36, "toolchain": 36, "aim": 36, "aggreg": 36, "autom": 36, "unifi": 36, "serv": [36, 43], "plugin": 36, "innov": 36, "e2": 36, "solut": 36, "organ": [36, 37, 40, 43], "overview": [36, 40], "introduct": 36, "guidanc": 36, "tutori": [36, 38, 40], "vscode": 37, "intellisens": 37, "yml": 37, "suppurt": 37, "authent": 37, "credenti": 37, "too": 37, "3000": 37, "sometim": [37, 43], "hang": 37, "stream": 37, "issu": 37, "fail": 37, "exclude_managed_identity_credenti": 37, "4000": 37, "item": [37, 42, 43], "insensit": [37, 46], "Its": 37, "hub": [37, 45], "equival": [37, 38], "causal": 37, "try": [37, 42], "index": 37, "string_to_int_dim_param": 37, "kv_cach": [37, 43], "ort_present_key_nam": 37, "ort_present_value_nam": 37, "world_siz": 37, "num_hidden_lay": 37, "num_attention_head": 37, "past_sequence_length": 37, "sequence_length_idx": 37, "past_kv_dynamic_axi": 37, "present_kv_dynamic_axi": 37, "from_pretrain": 37, "tabl": [37, 43], "alia": [37, 47], "uniqu": [37, 43], "among": 37, "torchmetr": 37, "equal": 37, "metric_config": [37, 39, 44], "alter": 37, "subtyp": 37, "accuracysubtyp": [37, 44], "latencysubtyp": [37, 39, 44], "throughputsubtyp": [37, 44], "load_param": 37, "word_length": 37, "compute_param": 37, "pred": [37, 44], "result_kei": 37, "9": 37, "boolean": [37, 47], "percent": [37, 42, 44], "evaluate_func_kwarg": [37, 44], "signatur": 37, "posit": [37, 43], "metric_func_kwarg": 37, "pars": 37, "our": 37, "cifar": 37, "resourcegroup": 37, "create_dataload": [37, 44], "20": [37, 42, 44], "definit": 37, "pass_flow": [37, 42], "decid": [37, 46], "overridden": [37, 46], "onnxtransformersoptim": 37, "snpetoonnxconvers": 37, "xilinx": 37, "On": 37, "optimumconvers": 37, "optimummerg": 37, "mixedprecisionoverrid": 37, "onnx_convers": [37, 38], "calib_data_coonfig": 37, "transformers_optim": 37, "768": 37, "qualifi": 37, "stop": 37, "met": 37, "plot": 37, "verbos": [37, 38], "warn": 37, "error": [37, 47], "fatal": 37, "ort_log_severity_level": 37, "ort_py_log_severity_level": 37, "focu": 38, "simplest": 38, "my_model_acceleration_descript": 38, "local_requir": 38, "remote_requir": 38, "olive_config": 38, "launch": 38, "temporari": 38, "reason": 38, "disk": 38, "permiss": 38, "cuda_visible_devic": 38, "know": 38, "term": 38, "slightli": 38, "pytorchmodel": [38, 39, 45], "user_provided_info": 39, "user_provided_info_json_fil": 39, "loader": 39, "parse_obj": 39, "metrictyp": [39, 44], "latency_metr": [39, 44], "warmup_num": 39, "repeat_test_num": 39, "sleep_num": 39, "latency_data_config": [39, 44], "engine_config": 39, "onnx_conversion_config": 39, "quantization_config": 39, "best_execut": 39, "default_cpu_acceler": 39, "k8": 40, "home": [40, 47], "kube": 40, "kubeconfig": 40, "lunch": 40, "infrastructur": 40, "minut": 40, "verif": 40, "attach": 40, "coupl": 41, "indic": [41, 43, 47], "mutual": 41, "aml_config": 41, "redund": [42, 46], "At": 42, "still": [42, 45], "stack": 42, "disable_auto_optim": 42, "bert_gpu": 42, "auto_optimizer_config": 42, "tensorrtexecutionprovid": [42, 47], "trt_fp16": 42, "flow": 42, "catch": 42, "conveni": 42, "comparison": 42, "uncas": [42, 43], "mrpc": [42, 43], "classif": [42, 43], "glue": [42, 43], "subset": [42, 43], "input_col": [42, 43], "sentence1": [42, 43], "sentence2": [42, 43], "label_col": [42, 43], "label": [42, 43, 47], "cuda_transformers_optim": 42, "trt_transformers_optim": 42, "cuda_perf_tun": 42, "trt_perf_tun": 42, "enable_trt_fp16": 42, "rule": 42, "expert": 42, "open_llama_sparsegpt_gpu": 43, "dataset_1": 43, "dataset_2": 43, "letter": 43, "underscor": 43, "common_pass": 43, "dive": 43, "dummydatacontain": 43, "dummy_data_config_templ": 43, "token_type_id": 43, "datacomponentconfig": [43, 44], "huggingface_data_config": 43, "rawdatacontain": 43, "raw_data": 43, "input_dir": 43, "input_suffix": 43, "input_order_fil": 43, "input_ord": 43, "transformersdummydatacontain": 43, "transformers_dummy_data_config": 43, "config_path": 43, "transformerspromptdummydatacontain": 43, "transformerstokendummydatacontain": 43, "datacontain": 43, "four": 43, "load_dataset": 43, "local_dataset": 43, "simple_dataset": 43, "huggingface_dataset": 43, "raw_dataset": 43, "pre_process_data": 43, "pre_process": 43, "huggingface_pre_process": 43, "ner_huggingface_preprocess": 43, "text_generation_huggingface_pre_process": 43, "post_process_data": 43, "text_classification_post_process": 43, "ner_post_process": 43, "text_generation_post_process": 43, "default_dataload": 43, "no_auto_batch_dataload": 43, "show": 43, "rewrit": 43, "But": 43, "user_dir": 43, "customized_huggingface_dataset": 43, "register_dataset": 43, "_output": 43, "thing": [43, 46], "notic": 43, "register_pre_process": 43, "customized_huggingface_pre_process": 43, "customized_post_process": 43, "customized_dataload": 43, "inception_post_process": 43, "l8": 43, "l10": 43, "json_config": 43, "inception_config": 43, "l14": 43, "l16": 43, "conflict": 43, "metricgo": 44, "accuracy_metr": 44, "throughput_data_config": 44, "throughput_metr": 44, "accuracy_custom": 44, "eval_accuraci": 44, "yourself": 44, "model_output": 44, "omit": [44, 45], "my_model_dir": 45, "curat": 45, "fulli": 46, "olive_pass": 46, "create_pass_from_dict": 46, "mandatori": 47, "rais": 47, "localtargetuserconfig": 47, "prepend": 47, "python_system": 47, "virtualenv": 47, "myenv": 47, "openvinoexecutionprovid": 47, "python_environment_system": 47, "dmlexecutionprovid": 47, "pythonenvironmenttargetuserconfig": 47, "docker_system": 47, "mnist_requir": 47, "count": 47, "propag": 47, "cuda11": 47, "cudnn8": 47, "ubuntu22": 47, "filter": 47, "therefor": 47, "azureml_system": 47, "azuremltargetuserconfig": 47, "relev": 47, "isolatedorttargetuserconfig": 47}, "objects": {"olive.azureml.azureml_client": [[5, 0, 1, "", "AzureMLClientConfig"]], "olive.azureml.azureml_client.AzureMLClientConfig": [[5, 1, 1, "", "aml_config_path"], [5, 2, 1, "", "create_client"], [5, 2, 1, "", "create_registry_client"], [5, 1, 1, "", "default_auth_params"], [5, 2, 1, "", "get_workspace_config"], [5, 1, 1, "", "keyvault_name"], [5, 1, 1, "", "max_operation_retries"], [5, 1, 1, "", "operation_retry_interval"], [5, 1, 1, "", "read_timeout"], [5, 1, 1, "", "resource_group"], [5, 1, 1, "", "subscription_id"], [5, 1, 1, "", "workspace_name"]], "olive.engine": [[0, 3, 1, "", "Engine"], [0, 0, 1, "", "EngineConfig"]], "olive.engine.Engine": [[0, 2, 1, "", "register"], [0, 2, 1, "", "run"]], "olive.engine.EngineConfig": [[0, 1, 1, "", "cache_dir"], [0, 1, 1, "", "clean_cache"], [0, 1, 1, "", "clean_evaluation_cache"], [0, 1, 1, "", "evaluator"], [0, 1, 1, "", "host"], [0, 1, 1, "", "plot_pareto_frontier"], [0, 1, 1, "", "search_strategy"], [0, 1, 1, "", "target"]], "olive.evaluator.metric": [[2, 3, 1, "", "AccuracySubType"], [2, 3, 1, "", "LatencySubType"], [2, 0, 1, "", "Metric"], [2, 0, 1, "", "MetricGoal"], [2, 3, 1, "", "MetricType"], [2, 3, 1, "", "ThroughputSubType"]], "olive.evaluator.metric.AccuracySubType": [[2, 4, 1, "", "ACCURACY_SCORE"], [2, 4, 1, "", "AUROC"], [2, 4, 1, "", "F1_SCORE"], [2, 4, 1, "", "PERPLEXITY"], [2, 4, 1, "", "PRECISION"], [2, 4, 1, "", "RECALL"]], "olive.evaluator.metric.LatencySubType": [[2, 4, 1, "", "AVG"], [2, 4, 1, "", "MAX"], [2, 4, 1, "", "MIN"], [2, 4, 1, "", "P50"], [2, 4, 1, "", "P75"], [2, 4, 1, "", "P90"], [2, 4, 1, "", "P95"], [2, 4, 1, "", "P99"], [2, 4, 1, "", "P999"]], "olive.evaluator.metric.Metric": [[2, 1, 1, "", "backend"], [2, 1, 1, "", "data_config"], [2, 2, 1, "", "get_inference_settings"], [2, 2, 1, "", "get_run_kwargs"], [2, 2, 1, "", "get_sub_type_info"], [2, 1, 1, "", "name"], [2, 1, 1, "", "sub_types"], [2, 1, 1, "", "type"], [2, 1, 1, "", "user_config"]], "olive.evaluator.metric.MetricGoal": [[2, 2, 1, "", "has_regression_goal"], [2, 1, 1, "", "type"], [2, 1, 1, "", "value"]], "olive.evaluator.metric.MetricType": [[2, 4, 1, "", "ACCURACY"], [2, 4, 1, "", "CUSTOM"], [2, 4, 1, "", "LATENCY"], [2, 4, 1, "", "THROUGHPUT"]], "olive.evaluator.metric.ThroughputSubType": [[2, 4, 1, "", "AVG"], [2, 4, 1, "", "MAX"], [2, 4, 1, "", "MIN"], [2, 4, 1, "", "P50"], [2, 4, 1, "", "P75"], [2, 4, 1, "", "P90"], [2, 4, 1, "", "P95"], [2, 4, 1, "", "P99"], [2, 4, 1, "", "P999"]], "olive.evaluator.olive_evaluator": [[1, 3, 1, "", "OliveEvaluator"], [0, 0, 1, "", "OliveEvaluatorConfig"]], "olive.evaluator.olive_evaluator.OliveEvaluatorConfig": [[0, 2, 1, "", "create_evaluator"], [0, 5, 1, "", "is_accuracy_drop_tolerance"], [0, 1, 1, "", "metrics"], [0, 1, 1, "", "script_dir"], [0, 1, 1, "", "type"], [0, 1, 1, "", "type_args"], [0, 1, 1, "", "user_script"]], "olive.hardware.accelerator": [[7, 3, 1, "", "Device"]], "olive.hardware.accelerator.Device": [[7, 4, 1, "", "CPU"], [7, 4, 1, "", "CPU_SPR"], [7, 4, 1, "", "GPU"], [7, 4, 1, "", "INTEL_MYRIAD"], [7, 4, 1, "", "NPU"], [7, 4, 1, "", "VPU"]], "olive.model": [[3, 3, 1, "", "CompositeModelHandler"], [3, 3, 1, "", "DistributedHfModelHandler"], [3, 3, 1, "", "DistributedOnnxModelHandler"], [3, 3, 1, "", "HfModelHandler"], [3, 3, 1, "", "ModelConfig"], [3, 3, 1, "", "ONNXModelHandler"], [3, 3, 1, "", "OpenVINOModelHandler"], [3, 3, 1, "", "PyTorchModelHandler"], [3, 3, 1, "", "SNPEModelHandler"]], "olive.passes.pytorch.lora": [[4, 0, 1, "", "HFTrainingArguments"]], "olive.passes.pytorch.lora.HFTrainingArguments": [[4, 2, 1, "", "create_training_args"], [4, 1, 1, "", "evaluation_strategy"], [4, 1, 1, "", "extra_args"], [4, 1, 1, "", "gradient_checkpointing"], [4, 1, 1, "", "learning_rate"], [4, 1, 1, "", "lr_scheduler_type"], [4, 1, 1, "", "optim"], [4, 1, 1, "", "output_dir"], [4, 1, 1, "", "overwrite_output_dir"], [4, 1, 1, "", "report_to"], [4, 1, 1, "", "resume_from_checkpoint"], [4, 1, 1, "", "warmup_ratio"]], "olive.resource_path": [[5, 3, 1, "", "ResourceType"]], "olive.resource_path.ResourceType": [[5, 4, 1, "", "AzureMLDatastore"], [5, 4, 1, "", "AzureMLJobOutput"], [5, 4, 1, "", "AzureMLModel"], [5, 4, 1, "", "AzureMLRegistryModel"], [5, 4, 1, "", "LocalFile"], [5, 4, 1, "", "LocalFolder"], [5, 4, 1, "", "StringName"]], "olive.strategy.search_strategy": [[0, 3, 1, "", "SearchStrategy"], [0, 0, 1, "", "SearchStrategyConfig"]], "olive.strategy.search_strategy.SearchStrategyConfig": [[0, 1, 1, "", "execution_order"], [0, 1, 1, "", "max_iter"], [0, 1, 1, "", "max_time"], [0, 1, 1, "", "output_model_num"], [0, 1, 1, "", "search_algorithm"], [0, 1, 1, "", "search_algorithm_config"], [0, 1, 1, "", "stop_when_goals_met"]], "olive.systems.azureml": [[7, 0, 1, "", "AzureMLDockerConfig"], [7, 3, 1, "", "AzureMLSystem"]], "olive.systems.azureml.AzureMLDockerConfig": [[7, 1, 1, "", "base_image"], [7, 1, 1, "", "build_context_path"], [7, 1, 1, "", "conda_file_path"], [7, 1, 1, "", "dockerfile"], [7, 1, 1, "", "name"], [7, 1, 1, "", "version"]], "olive.systems.common": [[7, 3, 1, "", "AcceleratorConfig"], [7, 3, 1, "", "SystemType"]], "olive.systems.common.AcceleratorConfig": [[7, 1, 1, "", "device"], [7, 1, 1, "", "execution_providers"]], "olive.systems.common.SystemType": [[7, 4, 1, "", "AzureML"], [7, 4, 1, "", "Docker"], [7, 4, 1, "", "IsolatedORT"], [7, 4, 1, "", "Local"], [7, 4, 1, "", "PythonEnvironment"]], "olive.systems.docker": [[7, 3, 1, "", "DockerSystem"], [7, 0, 1, "", "LocalDockerConfig"]], "olive.systems.docker.LocalDockerConfig": [[7, 1, 1, "", "build_args"], [7, 1, 1, "", "build_context_path"], [7, 1, 1, "", "dockerfile"], [7, 1, 1, "", "image_name"], [7, 1, 1, "", "run_params"]], "olive.systems.isolated_ort": [[7, 3, 1, "", "IsolatedORTSystem"]], "olive.systems.local": [[7, 3, 1, "", "LocalSystem"]], "olive.systems.python_environment": [[7, 3, 1, "", "PythonEnvironmentSystem"]], "olive.systems": [[7, 6, 0, "-", "system_alias"]], "olive.systems.system_alias": [[7, 3, 1, "", "AzureND12SSystem"], [7, 3, 1, "", "AzureND24RSSystem"], [7, 3, 1, "", "AzureND24SSystem"], [7, 3, 1, "", "AzureND6SSystem"], [7, 3, 1, "", "AzureND96A100System"], [7, 3, 1, "", "AzureND96ASystem"], [7, 3, 1, "", "AzureNDV2System"], [7, 3, 1, "", "SurfaceProSystem1796"], [7, 3, 1, "", "SurfaceProSystem1807"]], "olive.systems.system_alias.AzureND12SSystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND24RSSystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND24SSystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND6SSystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND96A100System": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND96ASystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureNDV2System": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.SurfaceProSystem1796": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.SurfaceProSystem1807": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_config": [[7, 0, 1, "", "AzureMLTargetUserConfig"], [7, 0, 1, "", "DockerTargetUserConfig"], [7, 0, 1, "", "IsolatedORTTargetUserConfig"], [7, 0, 1, "", "LocalTargetUserConfig"], [7, 0, 1, "", "PythonEnvironmentTargetUserConfig"], [7, 0, 1, "", "SystemConfig"]], "olive.systems.system_config.AzureMLTargetUserConfig": [[7, 1, 1, "", "aml_compute"], [7, 1, 1, "", "aml_docker_config"], [7, 1, 1, "", "aml_environment_config"], [7, 1, 1, "", "azureml_client_config"], [7, 1, 1, "", "datastores"], [7, 1, 1, "", "instance_count"], [7, 1, 1, "", "is_dev"], [7, 1, 1, "", "olive_managed_env"], [7, 1, 1, "", "requirements_file"], [7, 1, 1, "", "resources"], [7, 1, 1, "", "tags"]], "olive.systems.system_config.DockerTargetUserConfig": [[7, 1, 1, "", "is_dev"], [7, 1, 1, "", "local_docker_config"], [7, 1, 1, "", "olive_managed_env"], [7, 1, 1, "", "requirements_file"]], "olive.systems.system_config.IsolatedORTTargetUserConfig": [[7, 1, 1, "", "environment_variables"], [7, 1, 1, "", "prepend_to_path"], [7, 1, 1, "", "python_environment_path"]], "olive.systems.system_config.LocalTargetUserConfig": [[7, 1, 1, "", "accelerators"], [7, 1, 1, "", "hf_token"]], "olive.systems.system_config.PythonEnvironmentTargetUserConfig": [[7, 1, 1, "", "olive_managed_env"], [7, 1, 1, "", "requirements_file"]], "olive.systems.system_config.SystemConfig": [[7, 1, 1, "", "config"], [7, 2, 1, "", "create_system"], [7, 5, 1, "", "olive_managed_env"], [7, 1, 1, "", "type"]], "": [[4, 7, 1, "cmdoption-arg-ActivationSymmetric", "ActivationSymmetric"], [4, 7, 1, "cmdoption-arg-AddQDQPairToWeight", "AddQDQPairToWeight"], [4, 7, 1, "cmdoption-arg-EnableSubgraph", "EnableSubgraph"], [4, 7, 1, "cmdoption-arg-ForceQuantizeNoInputCheck", "ForceQuantizeNoInputCheck"], [4, 7, 1, "cmdoption-arg-MatMulConstBOnly", "MatMulConstBOnly"], [4, 7, 1, "cmdoption-arg-WeightSymmetric", "WeightSymmetric"], [4, 7, 1, "cmdoption-arg-a_asym", "a_asym"], [4, 7, 1, "cmdoption-arg-a_bits", "a_bits"], [4, 7, 1, "cmdoption-arg-a_clip_ratio", "a_clip_ratio"], [4, 7, 1, "cmdoption-arg-a_groupsize", "a_groupsize"], [4, 7, 1, "cmdoption-arg-activation_type", "activation_type"], [4, 7, 1, "cmdoption-arg-all_tensors_to_one_file", "all_tensors_to_one_file"], [4, 7, 1, "cmdoption-arg-allow_tf32", "allow_tf32"], [4, 7, 1, "cmdoption-arg-append_first_op_types_to_quantize_list", "append_first_op_types_to_quantize_list"], [4, 7, 1, "cmdoption-arg-approach", "approach"], [4, 7, 1, "cmdoption-arg-atol", "atol"], [5, 7, 1, "cmdoption-arg-azureml_client", "azureml_client"], [4, 7, 1, "cmdoption-arg-backend", "backend"], [4, 7, 1, "cmdoption-arg-binary_file", "binary_file"], [4, 7, 1, "cmdoption-arg-bits", "bits"], [4, 7, 1, "cmdoption-arg-blocksize", "blocksize"], [4, 7, 1, "cmdoption-arg-calibrate_method", "calibrate_method"], [4, 7, 1, "cmdoption-arg-calibration_batch_size", "calibration_batch_size"], [4, 7, 1, "cmdoption-arg-calibration_data_config", "calibration_data_config"], [4, 7, 1, "cmdoption-arg-calibration_nsamples", "calibration_nsamples"], [4, 7, 1, "cmdoption-arg-calibration_sampling_size", "calibration_sampling_size"], [4, 7, 1, "cmdoption-arg-checkpoint_path", "checkpoint_path"], [4, 7, 1, "cmdoption-arg-components", "components"], [4, 7, 1, "cmdoption-arg-compress_to_fp16", "compress_to_fp16"], [4, 7, 1, "cmdoption-arg-compute_dtype", "compute_dtype"], [4, 7, 1, "cmdoption-arg-convert_attribute", "convert_attribute"], [4, 7, 1, "cmdoption-arg-cpu_cores", "cpu_cores"], [4, 7, 1, "cmdoption-arg-damp_percent", "damp_percent"], [4, 7, 1, "cmdoption-arg-data_config", "data_config"], [5, 7, 1, "cmdoption-arg-datastore_name", "datastore_name"], [5, 7, 1, "cmdoption-arg-datastore_url", "datastore_url"], [4, 7, 1, "cmdoption-arg-desc_act", "desc_act"], [4, 7, 1, "cmdoption-arg-device", "device"], [4, 7, 1, "cmdoption-arg-dim_param", "dim_param"], [4, 7, 1, "cmdoption-arg-dim_value", "dim_value"], [4, 7, 1, "cmdoption-arg-do_validate", "do_validate"], [4, 7, 1, "cmdoption-arg-domain", "domain"], [4, 7, 1, "cmdoption-arg-double_quant", "double_quant"], [4, 7, 1, "cmdoption-arg-duo_scaling", "duo_scaling"], [4, 7, 1, "cmdoption-arg-dynamic_lora_r", "dynamic_lora_r"], [4, 7, 1, "cmdoption-arg-element_wise_binary_ops", "element_wise_binary_ops"], [4, 7, 1, "cmdoption-arg-enable_cuda_graph", "enable_cuda_graph"], [4, 7, 1, "cmdoption-arg-enable_dpu", "enable_dpu"], [4, 7, 1, "cmdoption-arg-enable_htp", "enable_htp"], [4, 7, 1, "cmdoption-arg-enable_profiling", "enable_profiling"], [4, 7, 1, "cmdoption-arg-eval_data_config", "eval_data_config"], [4, 7, 1, "cmdoption-arg-example_input_func", "example_input_func"], [4, 7, 1, "cmdoption-arg-exclude_embeds", "exclude_embeds"], [4, 7, 1, "cmdoption-arg-exclude_lm_head", "exclude_lm_head"], [4, 7, 1, "cmdoption-arg-excluded_precisions", "excluded_precisions"], [4, 7, 1, "cmdoption-arg-execution_mode_list", "execution_mode_list"], [4, 7, 1, "cmdoption-arg-export_compatible", "export_compatible"], [4, 7, 1, "cmdoption-arg-external_data_name", "external_data_name"], [4, 7, 1, "cmdoption-arg-extra_args", "extra_args"], [4, 7, 1, "cmdoption-arg-extra_configs", "extra_configs"], [4, 7, 1, "cmdoption-arg-extra_options", "extra_options"], [4, 7, 1, "cmdoption-arg-extra_session_config", "extra_session_config"], [4, 7, 1, "cmdoption-arg-final_orientation", "final_orientation"], [4, 7, 1, "cmdoption-arg-float16", "float16"], [4, 7, 1, "cmdoption-arg-force_evaluate_other_eps", "force_evaluate_other_eps"], [4, 7, 1, "cmdoption-arg-force_fp16_inputs", "force_fp16_inputs"], [4, 7, 1, "cmdoption-arg-force_fp32_nodes", "force_fp32_nodes"], [4, 7, 1, "cmdoption-arg-force_fp32_ops", "force_fp32_ops"], [4, 7, 1, "cmdoption-arg-fp16", "fp16"], [4, 7, 1, "cmdoption-arg-fuse_layernorm", "fuse_layernorm"], [4, 7, 1, "cmdoption-arg-gptq_damping", "gptq_damping"], [4, 7, 1, "cmdoption-arg-gptq_opt_scales", "gptq_opt_scales"], [4, 7, 1, "cmdoption-arg-gpus", "gpus"], [6, 7, 1, "cmdoption-arg-group", "group"], [4, 7, 1, "cmdoption-arg-group_size", "group_size"], [4, 7, 1, "cmdoption-arg-hidden_size", "hidden_size"], [4, 7, 1, "cmdoption-arg-htp_socs", "htp_socs"], [4, 7, 1, "cmdoption-arg-ignored_scope", "ignored_scope"], [4, 7, 1, "cmdoption-arg-ignored_scope_type", "ignored_scope_type"], [4, 7, 1, "cmdoption-arg-input", "input"], [4, 7, 1, "cmdoption-arg-input_dim", "input_dim"], [4, 7, 1, "cmdoption-arg-input_int32", "input_int32"], [4, 7, 1, "cmdoption-arg-input_layouts", "input_layouts"], [4, 7, 1, "cmdoption-arg-input_model_dtype", "input_model_dtype"], [4, 7, 1, "cmdoption-arg-input_name", "input_name"], [4, 7, 1, "cmdoption-arg-input_names", "input_names"], [4, 7, 1, "cmdoption-arg-input_nodes", "input_nodes"], [4, 7, 1, "cmdoption-arg-input_shape", "input_shape"], [4, 7, 1, "cmdoption-arg-input_shapes", "input_shapes"], [4, 7, 1, "cmdoption-arg-input_types", "input_types"], [4, 7, 1, "cmdoption-arg-inputs_to_make_channel_last", "inputs_to_make_channel_last"], [4, 7, 1, "cmdoption-arg-inside_layer_modules", "inside_layer_modules"], [4, 7, 1, "cmdoption-arg-int4_accuracy_level", "int4_accuracy_level"], [4, 7, 1, "cmdoption-arg-int4_block_size", "int4_block_size"], [4, 7, 1, "cmdoption-arg-inter_thread_num_list", "inter_thread_num_list"], [4, 7, 1, "cmdoption-arg-intra_thread_num_list", "intra_thread_num_list"], [4, 7, 1, "cmdoption-arg-io_bind", "io_bind"], [5, 7, 1, "cmdoption-arg-job_name", "job_name"], [4, 7, 1, "cmdoption-arg-k_bits", "k_bits"], [4, 7, 1, "cmdoption-arg-k_clip_ratio", "k_clip_ratio"], [4, 7, 1, "cmdoption-arg-k_groupsize", "k_groupsize"], [4, 7, 1, "cmdoption-arg-keep_io_types", "keep_io_types"], [4, 7, 1, "cmdoption-arg-layer_name_filter", "layer_name_filter"], [4, 7, 1, "cmdoption-arg-layers_block_name", "layers_block_name"], [4, 7, 1, "cmdoption-arg-lib_name", "lib_name"], [4, 7, 1, "cmdoption-arg-lib_targets", "lib_targets"], [4, 7, 1, "cmdoption-arg-loftq_iter", "loftq_iter"], [4, 7, 1, "cmdoption-arg-logger", "logger"], [4, 7, 1, "cmdoption-arg-lora_alpha", "lora_alpha"], [4, 7, 1, "cmdoption-arg-lora_dropout", "lora_dropout"], [4, 7, 1, "cmdoption-arg-lora_r", "lora_r"], [4, 7, 1, "cmdoption-arg-make_inputs", "make_inputs"], [4, 7, 1, "cmdoption-arg-max_finite_val", "max_finite_val"], [4, 7, 1, "cmdoption-arg-max_layer", "max_layer"], [4, 7, 1, "cmdoption-arg-merge_adapter_weights", "merge_adapter_weights"], [4, 7, 1, "cmdoption-arg-metadata_only", "metadata_only"], [4, 7, 1, "cmdoption-arg-metric", "metric"], [4, 7, 1, "cmdoption-arg-min_layer", "min_layer"], [4, 7, 1, "cmdoption-arg-min_positive_val", "min_positive_val"], [4, 7, 1, "cmdoption-arg-model_type", "model_type"], [4, 7, 1, "cmdoption-arg-modules_to_fuse", "modules_to_fuse"], [4, 7, 1, "cmdoption-arg-modules_to_not_convert", "modules_to_not_convert"], [4, 7, 1, "cmdoption-arg-modules_to_save", "modules_to_save"], [6, 7, 1, "cmdoption-arg-multivariate", "multivariate"], [5, 7, 1, "cmdoption-arg-name", "name"], [4, 7, 1, "cmdoption-arg-name_pattern", "name_pattern"], [4, 7, 1, "cmdoption-arg-need_layer_fusing", "need_layer_fusing"], [4, 7, 1, "cmdoption-arg-no_repeat_ngram_size", "no_repeat_ngram_size"], [4, 7, 1, "cmdoption-arg-node_block_list", "node_block_list"], [4, 7, 1, "cmdoption-arg-nodes_to_exclude", "nodes_to_exclude"], [4, 7, 1, "cmdoption-arg-nodes_to_quantize", "nodes_to_quantize"], [4, 7, 1, "cmdoption-arg-nsamples", "nsamples"], [4, 7, 1, "cmdoption-arg-num_epochs", "num_epochs"], [4, 7, 1, "cmdoption-arg-num_heads", "num_heads"], [4, 7, 1, "cmdoption-arg-num_key_value_heads", "num_key_value_heads"], [6, 7, 1, "cmdoption-arg-num_samples", "num_samples"], [4, 7, 1, "cmdoption-arg-num_steps", "num_steps"], [4, 7, 1, "cmdoption-arg-only_onnxruntime", "only_onnxruntime"], [4, 7, 1, "cmdoption-arg-op_block_list", "op_block_list"], [4, 7, 1, "cmdoption-arg-op_type_dict", "op_type_dict"], [4, 7, 1, "cmdoption-arg-op_types_to_quantize", "op_types_to_quantize"], [4, 7, 1, "cmdoption-arg-opt_level", "opt_level"], [4, 7, 1, "cmdoption-arg-opt_level_list", "opt_level_list"], [4, 7, 1, "cmdoption-arg-optimization_options", "optimization_options"], [4, 7, 1, "cmdoption-arg-optimize_model", "optimize_model"], [4, 7, 1, "cmdoption-arg-optional_inputs", "optional_inputs"], [4, 7, 1, "cmdoption-arg-out_node", "out_node"], [4, 7, 1, "cmdoption-arg-output_model", "output_model"], [5, 7, 1, "cmdoption-arg-output_name", "output_name"], [4, 7, 1, "cmdoption-arg-output_names", "output_names"], [4, 7, 1, "cmdoption-arg-output_nodes", "output_nodes"], [4, 7, 1, "cmdoption-arg-outputs_to_make_channel_last", "outputs_to_make_channel_last"], [4, 7, 1, "cmdoption-arg-outside_layer_modules", "outside_layer_modules"], [4, 7, 1, "cmdoption-arg-overrides_config", "overrides_config"], [4, 7, 1, "cmdoption-arg-pack_inputs", "pack_inputs"], [4, 7, 1, "cmdoption-arg-pack_model_for_onnx_conversion", "pack_model_for_onnx_conversion"], [4, 7, 1, "cmdoption-arg-parallel_jobs", "parallel_jobs"], [4, 7, 1, "cmdoption-arg-past_key_value_name", "past_key_value_name"], [5, 7, 1, "cmdoption-arg-path", "path"], [4, 7, 1, "cmdoption-arg-per_channel", "per_channel"], [4, 7, 1, "cmdoption-arg-percdamp", "percdamp"], [4, 7, 1, "cmdoption-arg-post", "post"], [4, 7, 1, "cmdoption-arg-pre", "pre"], [4, 7, 1, "cmdoption-arg-precision", "precision"], [4, 7, 1, "cmdoption-arg-prepare_qnn_config", "prepare_qnn_config"], [4, 7, 1, "cmdoption-arg-preset", "preset"], [4, 7, 1, "cmdoption-arg-providers_list", "providers_list"], [4, 7, 1, "cmdoption-arg-ptl_data_module", "ptl_data_module"], [4, 7, 1, "cmdoption-arg-ptl_module", "ptl_module"], [4, 7, 1, "cmdoption-arg-q_group_size", "q_group_size"], [4, 7, 1, "cmdoption-arg-qconfig_func", "qconfig_func"], [4, 7, 1, "cmdoption-arg-qnn_extra_options", "qnn_extra_options"], [4, 7, 1, "cmdoption-arg-quant_format", "quant_format"], [4, 7, 1, "cmdoption-arg-quant_level", "quant_level"], [4, 7, 1, "cmdoption-arg-quant_mode", "quant_mode"], [4, 7, 1, "cmdoption-arg-quant_preprocess", "quant_preprocess"], [4, 7, 1, "cmdoption-arg-quant_type", "quant_type"], [4, 7, 1, "cmdoption-arg-recipes", "recipes"], [4, 7, 1, "cmdoption-arg-reduce_range", "reduce_range"], [5, 7, 1, "cmdoption-arg-relative_path", "relative_path"], [4, 7, 1, "cmdoption-arg-rotate", "rotate"], [4, 7, 1, "cmdoption-arg-rotation_seed", "rotation_seed"], [4, 7, 1, "cmdoption-arg-round_interval", "round_interval"], [4, 7, 1, "cmdoption-arg-save_as_external_data", "save_as_external_data"], [4, 7, 1, "cmdoption-arg-save_format", "save_format"], [4, 7, 1, "cmdoption-arg-save_metadata_for_token_generation", "save_metadata_for_token_generation"], [4, 7, 1, "cmdoption-arg-script_dir", "script_dir"], [4, 7, 1, "cmdoption-arg-search", "search"], [4, 7, 1, "cmdoption-arg-seed", "seed"], [4, 7, 1, "cmdoption-arg-size_threshold", "size_threshold"], [4, 7, 1, "cmdoption-arg-sparsity", "sparsity"], [4, 7, 1, "cmdoption-arg-static_groups", "static_groups"], [4, 7, 1, "cmdoption-arg-strict", "strict"], [4, 7, 1, "cmdoption-arg-sym", "sym"], [4, 7, 1, "cmdoption-arg-target_device", "target_device"], [4, 7, 1, "cmdoption-arg-target_modules", "target_modules"], [4, 7, 1, "cmdoption-arg-target_opset", "target_opset"], [4, 7, 1, "cmdoption-arg-tool_command", "tool_command"], [4, 7, 1, "cmdoption-arg-tool_command_args", "tool_command_args"], [4, 7, 1, "cmdoption-arg-torch_dtype", "torch_dtype"], [4, 7, 1, "cmdoption-arg-train_data_config", "train_data_config"], [4, 7, 1, "cmdoption-arg-training_args", "training_args"], [4, 7, 1, "cmdoption-arg-training_loop_func", "training_loop_func"], [4, 7, 1, "cmdoption-arg-trt_fp16_enable", "trt_fp16_enable"], [4, 7, 1, "cmdoption-arg-true_sequential", "true_sequential"], [4, 7, 1, "cmdoption-arg-tuning_criterion", "tuning_criterion"], [4, 7, 1, "cmdoption-arg-use_dynamo_exporter", "use_dynamo_exporter"], [4, 7, 1, "cmdoption-arg-use_enhanced_quantizer", "use_enhanced_quantizer"], [4, 7, 1, "cmdoption-arg-use_external_data_format", "use_external_data_format"], [4, 7, 1, "cmdoption-arg-use_forced_decoder_ids", "use_forced_decoder_ids"], [4, 7, 1, "cmdoption-arg-use_gpu", "use_gpu"], [4, 7, 1, "cmdoption-arg-use_gqa", "use_gqa"], [4, 7, 1, "cmdoption-arg-use_logits_processor", "use_logits_processor"], [4, 7, 1, "cmdoption-arg-use_prefix_vocab_mask", "use_prefix_vocab_mask"], [4, 7, 1, "cmdoption-arg-use_symbolic_shape_infer", "use_symbolic_shape_infer"], [4, 7, 1, "cmdoption-arg-use_temperature", "use_temperature"], [4, 7, 1, "cmdoption-arg-use_vocab_mask", "use_vocab_mask"], [4, 7, 1, "cmdoption-arg-user_script", "user_script"], [4, 7, 1, "cmdoption-arg-v_bits", "v_bits"], [4, 7, 1, "cmdoption-arg-v_clip_ratio", "v_clip_ratio"], [4, 7, 1, "cmdoption-arg-v_groupsize", "v_groupsize"], [4, 7, 1, "cmdoption-arg-val_data_config", "val_data_config"], [4, 7, 1, "cmdoption-arg-version", "version"], [4, 7, 1, "cmdoption-arg-w_asym", "w_asym"], [4, 7, 1, "cmdoption-arg-w_bit", "w_bit"], [4, 7, 1, "cmdoption-arg-w_bits", "w_bits"], [4, 7, 1, "cmdoption-arg-w_gptq", "w_gptq"], [4, 7, 1, "cmdoption-arg-w_groupsize", "w_groupsize"], [4, 7, 1, "cmdoption-arg-w_rtn", "w_rtn"], [4, 7, 1, "cmdoption-arg-weight_only_config", "weight_only_config"], [4, 7, 1, "cmdoption-arg-weight_type", "weight_type"], [6, 7, 1, "cmdoption-arg-with_replacement", "with_replacement"], [4, 7, 1, "cmdoption-arg-workspace", "workspace"], [4, 7, 1, "cmdoption-arg-zero_point", "zero_point"]], "extra.Sigmoid": [[4, 7, 1, "cmdoption-arg-extra.Sigmoid.nnapi", "nnapi"]]}, "objtypes": {"0": "py:pydantic_settings", "1": "py:pydantic_field", "2": "py:method", "3": "py:class", "4": "py:attribute", "5": "py:property", "6": "py:module", "7": "std:cmdoption"}, "objnames": {"0": ["py", "pydantic_settings", "Python settings"], "1": ["py", "pydantic_field", "Python field"], "2": ["py", "method", "Python method"], "3": ["py", "class", "Python class"], "4": ["py", "attribute", "Python attribute"], "5": ["py", "property", "Python property"], "6": ["py", "module", "Python module"], "7": ["std", "cmdoption", "program option"]}, "titleterms": {"engin": [0, 10, 37, 38, 39], "engineconfig": 0, "searchstrategi": 0, "oliveevalu": 1, "metric": [2, 16, 19, 44], "metrictyp": 2, "accuracysubtyp": 2, "latencysubtyp": 2, "throughputsubtyp": 2, "metricgo": 2, "olivemodel": 3, "model": [3, 12, 13, 14, 15, 16, 18, 19, 20, 22, 23, 25, 28, 29, 32, 34, 37, 38, 39, 45], "configur": [3, 11, 13, 14, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 42, 43, 44, 46, 47], "hf": 3, "handler": 3, "distribut": 3, "pytorch": [3, 21, 24, 27], "onnx": [3, 20, 22, 26], "openvino": [3, 23], "snpe": [3, 28], "composit": 3, "pass": [4, 10, 11, 13, 37, 38, 39, 46], "onnxconvers": 4, "onnxopversionconvers": 4, "onnxmodeloptim": 4, "orttransformersoptim": 4, "ortperftun": 4, "onnxfloattofloat16": 4, "onnxiofloat16tofloat32": 4, "ortmixedprecis": 4, "qnnpreprocess": 4, "mixedprecisionoverrid": 4, "onnxdynamicquant": 4, "onnxstaticquant": 4, "onnxquant": 4, "dynamictofixedshap": 4, "incdynamicquant": 4, "incstaticquant": 4, "incquant": 4, "vitisaiquant": 4, "appendprepostprocessingop": 4, "insertbeamsearch": 4, "extractadapt": 4, "lora": [4, 13, 17, 24], "qlora": [4, 24], "loftq": [4, 24], "hftrainingargu": 4, "quantizationawaretrain": 4, "openvinoconvers": 4, "openvinoquant": 4, "snpeconvers": 4, "snpequant": 4, "snpetoonnxconvers": 4, "qnnconvers": 4, "qnnmodellibgener": 4, "qnncontextbinarygener": 4, "mergeadapterweight": [4, 24], "sparsegpt": [4, 24], "slicegpt": [4, 24], "quarot": [4, 27], "gptqquantiz": 4, "autoawqquant": 4, "torchtrtconvers": [4, 21], "optimumconvers": 4, "optimummerg": 4, "modelbuild": 4, "resourcepath": 5, "resourcetyp": 5, "localfil": 5, "localfold": 5, "stringnam": 5, "azuremlmodel": [5, 19], "azuremldatastor": 5, "azuremljoboutput": 5, "searchalgorithm": 6, "exhaustivesearchalgorithm": 6, "randomsearchalgorithm": 6, "tpesearchalgorithm": 6, "olivesystem": 7, "config": [7, 13, 16, 19, 32, 34, 37, 43], "systemtyp": 7, "devic": [7, 12], "acceleratorconfig": [7, 47], "localtargetuserconfig": 7, "dockertargetuserconfig": 7, "azuremltargetuserconfig": 7, "pythonenvironmenttargetuserconfig": 7, "isolatedorttargetuserconfig": 7, "class": [7, 11], "localsystem": 7, "azuremlsystem": 7, "dockersystem": 7, "pythonenvironmentsystem": 7, "isolatedortsystem": 7, "system": [7, 10, 16, 30, 37, 39, 47], "alia": 7, "exampl": [8, 9, 11, 16, 20, 21, 22, 23, 24, 25, 26, 27, 28, 33, 35, 37], "custom": [9, 43, 44], "script": [9, 12, 41], "user_script": 9, "script_dir": 9, "design": 10, "evalu": [10, 37, 39], "search": [10, 22], "strategi": 10, "execut": 10, "order": 10, "algorithm": 10, "how": [11, 14, 19, 42, 43, 44, 45, 46, 47], "add": [11, 40], "new": 11, "optim": [11, 18, 22, 32, 34, 42], "1": [11, 23], "defin": 11, "2": [11, 23], "3": 11, "implement": 11, "run": [11, 13, 30, 32, 34, 39], "function": 11, "azur": [12, 14, 16, 30, 37, 40, 41], "ml": [12, 16, 37, 41], "integr": [12, 16], "machin": [12, 30, 40], "learn": [12, 30, 40], "client": [12, 37], "oliv": [12, 13, 19, 23, 25, 28, 30, 35, 36, 37, 38], "core": 12, "us": [12, 14, 32, 34], "azureml": [12, 16, 30, 37, 45, 47], "regist": [12, 39], "curat": 12, "store": 12, "datastor": [12, 45], "from": [12, 23, 31], "an": [12, 30], "job": [12, 45], "output": [12, 20, 30, 45], "data": [12, 13, 34, 43], "comput": [12, 13, 30], "host": [12, 37, 39, 40], "target": [12, 39], "packag": [12, 19], "workspac": [12, 30], "connect": 12, "your": 12, "own": 12, "arc": [12, 40], "helper": 12, "command": 13, "line": 13, "tool": [13, 23], "sub": 13, "name": [13, 45], "argument": 13, "finetun": 13, "option": [13, 23, 31, 37], "dataset": [13, 16], "remot": [13, 30, 45], "export": 13, "adapt": [13, 17, 22], "int4": 13, "quantiz": [13, 23, 24, 25, 26, 28, 29], "qualcomm": 13, "sdk": [13, 25, 28], "manag": [13, 47], "aml": [13, 40], "tune": [13, 22], "session": 13, "param": 13, "which": 13, "mutual": 13, "exclus": 13, "huggingfac": [13, 16], "ar": [13, 32], "provid": 13, "user": [13, 39], "should": 13, "follow": 13, "modifi": 13, "default": 13, "pleas": 13, "refer": [13, 35], "contain": 13, "transformerstokendummydatacontain": 13, "more": 13, "detail": 13, "cloud": [13, 14], "cach": [13, 14], "what": [14, 19], "i": [14, 19], "instal": [14, 23, 30, 31], "extra": [14, 23, 30], "depend": [14, 30, 31], "blob": 14, "storag": 14, "note": 14, "convers": [15, 20, 23, 25, 28], "introduct": 16, "input": [16, 20, 37, 38, 39], "hub": 16, "local": [16, 40, 45, 47], "load": 16, "login": 16, "docker": [16, 47], "python": [16, 47], "environ": [16, 47], "e2": 16, "transform": [18, 22], "artifact": 19, "zipfil": 19, "candidatemodel": 19, "rank": 19, "json": [19, 32, 37], "file": [19, 45], "azuremldata": 19, "azuremldeploy": 19, "dockerfil": 19, "infer": 19, "float16": 20, "float32": 20, "mix": 20, "precis": 20, "convert": 20, "dynam": 20, "shape": 20, "fix": 20, "ort": [22, 47], "append": 22, "pre": 22, "post": [22, 23, 28], "process": 22, "op": 22, "insert": 22, "beam": 22, "perform": 22, "extract": 22, "prerequisit": [23, 25, 28, 32, 34, 47], "runtim": 23, "develop": 23, "pypi": 23, "train": [23, 24, 28], "ptq": [23, 28, 32, 34], "awar": 24, "qnn": 25, "download": [25, 28], "unzip": [25, 28], "librari": 25, "gener": [25, 43], "context": 25, "binari": 25, "onnxruntim": 26, "intel": 26, "neural": 26, "compressor": 26, "amd": 26, "viti": 26, "ai": 26, "autogptq": 27, "autoawq": 27, "workflow": [30, 37], "pip": [31, 32, 34], "sourc": 31, "edit": 31, "bert": 32, "cpu": [32, 34], "requir": [32, 34], "sampl": [32, 34], "The": 32, "techniqu": 32, "specifi": 32, "bert_ptq_cpu": 32, "automat": 32, "without": 32, "select": 32, "ani": 32, "quickstart": 33, "resnet": 34, "prepar": 34, "welcom": 35, "": 35, "document": 35, "overview": 35, "get": 35, "start": 35, "featur": 35, "extend": 35, "tutori": 35, "api": 35, "content": 36, "id": 37, "azureml_cli": 37, "aml_config_path": 37, "aml_config": 37, "field": 37, "inform": [37, 38], "flow": 37, "quick": 38, "tour": [38, 39], "need": 38, "acceler": [38, 47], "appli": 38, "advanc": 39, "self": 40, "kubernet": 40, "cluster": 40, "creat": 40, "extens": 40, "link": 40, "list": 41, "manage_compute_inst": 41, "usag": 41, "To": [42, 43, 44, 45, 46, 47], "auto": 42, "support": 43, "templat": 43, "titl": 43, "built": 43, "compon": 43, "type": 44, "accuraci": 44, "latenc": 44, "throughput": 44, "multi": 44, "set": 45, "path": 45, "folder": 45, "string": 45, "registri": 45, "nativ": 47, "readymad": 47, "isol": 47}, "envversion": {"sphinx.domains.c": 3, "sphinx.domains.changeset": 1, "sphinx.domains.citation": 1, "sphinx.domains.cpp": 9, "sphinx.domains.index": 1, "sphinx.domains.javascript": 3, "sphinx.domains.math": 2, "sphinx.domains.python": 4, "sphinx.domains.rst": 2, "sphinx.domains.std": 2, "sphinx.ext.intersphinx": 1, "sphinx.ext.todo": 2, "sphinx.ext.viewcode": 1, "sphinx": 58}, "alltitles": {"Engine": [[0, "engine"], [0, "id2"], [10, "engine"], [39, "engine"]], "EngineConfig": [[0, "engineconfig"]], "SearchStrategy": [[0, "searchstrategy"]], "OliveEvaluator": [[1, "oliveevaluator"], [1, "id1"]], "Metric": [[2, "metric"], [2, "id2"]], "MetricType": [[2, "metrictype"]], "AccuracySubType": [[2, "accuracysubtype"]], "LatencySubType": [[2, "latencysubtype"]], "ThroughputSubType": [[2, "throughputsubtype"]], "MetricGoal": [[2, "metricgoal"]], "OliveModels": [[3, "olivemodels"]], "Model Configuration": [[3, "model-configuration"]], "Hf Model Handler": [[3, "hf-model-handler"]], "Distributed Hf Model Handler": [[3, "distributed-hf-model-handler"]], "PyTorch Model Handler": [[3, "pytorch-model-handler"]], "ONNX Model Handler": [[3, "onnx-model-handler"]], "Distributed Onnx Model Handler": [[3, "distributed-onnx-model-handler"]], "OpenVINO Model Handler": [[3, "openvino-model-handler"]], "SNPE Model Handler": [[3, "snpe-model-handler"]], "Composite Model Handler": [[3, "composite-model-handler"]], "Passes": [[4, "passes"]], "OnnxConversion": [[4, "onnxconversion"]], "OnnxOpVersionConversion": [[4, "onnxopversionconversion"]], "OnnxModelOptimizer": [[4, "onnxmodeloptimizer"]], "OrtTransformersOptimization": [[4, "orttransformersoptimization"]], "OrtPerfTuning": [[4, "ortperftuning"]], "OnnxFloatToFloat16": [[4, "onnxfloattofloat16"]], "OnnxIOFloat16ToFloat32": [[4, "onnxiofloat16tofloat32"]], "OrtMixedPrecision": [[4, "ortmixedprecision"]], "QNNPreprocess": [[4, "qnnpreprocess"]], "MixedPrecisionOverrides": [[4, "mixedprecisionoverrides"]], "OnnxDynamicQuantization": [[4, "onnxdynamicquantization"]], "OnnxStaticQuantization": [[4, "onnxstaticquantization"]], "OnnxQuantization": [[4, "onnxquantization"]], "DynamicToFixedShape": [[4, "dynamictofixedshape"]], "IncDynamicQuantization": [[4, "incdynamicquantization"]], "IncStaticQuantization": [[4, "incstaticquantization"]], "IncQuantization": [[4, "incquantization"]], "VitisAIQuantization": [[4, "vitisaiquantization"]], "AppendPrePostProcessingOps": [[4, "appendprepostprocessingops"]], "InsertBeamSearch": [[4, "insertbeamsearch"]], "ExtractAdapters": [[4, "extractadapters"]], "LoRA": [[4, "lora"], [24, "lora"]], "QLoRA": [[4, "qlora"], [24, "qlora"]], "LoftQ": [[4, "loftq"], [24, "loftq"]], "LoRA/QLoRA/LoftQ HFTrainingArguments": [[4, "lora-qlora-loftq-hftrainingarguments"]], "QuantizationAwareTraining": [[4, "quantizationawaretraining"]], "OpenVINOConversion": [[4, "openvinoconversion"]], "OpenVINOQuantization": [[4, "openvinoquantization"]], "SNPEConversion": [[4, "snpeconversion"]], "SNPEQuantization": [[4, "snpequantization"]], "SNPEtoONNXConversion": [[4, "snpetoonnxconversion"]], "QNNConversion": [[4, "qnnconversion"]], "QNNModelLibGenerator": [[4, "qnnmodellibgenerator"]], "QNNContextBinaryGenerator": [[4, "qnncontextbinarygenerator"]], "MergeAdapterWeights": [[4, "mergeadapterweights"], [24, "mergeadapterweights"]], "SparseGPT": [[4, "sparsegpt"], [24, "sparsegpt"]], "SliceGPT": [[4, "slicegpt"], [24, "slicegpt"]], "QuaRot": [[4, "quarot"], [27, "quarot"]], "GptqQuantizer": [[4, "gptqquantizer"]], "AutoAWQQuantizer": [[4, "autoawqquantizer"]], "TorchTRTConversion": [[4, "torchtrtconversion"], [21, "torchtrtconversion"]], "OptimumConversion": [[4, "optimumconversion"]], "OptimumMerging": [[4, "optimummerging"]], "ModelBuilder": [[4, "modelbuilder"]], "ResourcePath": [[5, "resourcepath"]], "ResourceType": [[5, "resourcetype"]], "LocalFile": [[5, "localfile"]], "LocalFolder": [[5, "localfolder"]], "StringName": [[5, "stringname"]], "AzureMLModel": [[5, "azuremlmodel"]], "AzureMLDatastore": [[5, "azuremldatastore"]], "AzureMLJobOutput": [[5, "azuremljoboutput"]], "SearchAlgorithms": [[6, "searchalgorithms"]], "ExhaustiveSearchAlgorithm": [[6, "exhaustivesearchalgorithm"]], "RandomSearchAlgorithm": [[6, "randomsearchalgorithm"]], "TPESearchAlgorithm": [[6, "tpesearchalgorithm"]], "OliveSystems": [[7, "olivesystems"]], "Config": [[7, "config"]], "SystemType": [[7, "systemtype"]], "Device": [[7, "device"]], "AcceleratorConfig": [[7, "acceleratorconfig"], [47, null]], "LocalTargetUserConfig": [[7, "localtargetuserconfig"]], "DockerTargetUserConfig": [[7, "dockertargetuserconfig"]], "AzureMLTargetUserConfig": [[7, "azuremltargetuserconfig"]], "PythonEnvironmentTargetUserConfig": [[7, "pythonenvironmenttargetuserconfig"]], "IsolatedORTTargetUserConfig": [[7, "isolatedorttargetuserconfig"]], "Classes": [[7, "classes"]], "LocalSystem": [[7, "localsystem"]], "AzureMLSystem": [[7, "azuremlsystem"]], "DockerSystem": [[7, "dockersystem"]], "PythonEnvironmentSystem": [[7, "pythonenvironmentsystem"]], "IsolatedORTSystem": [[7, "isolatedortsystem"]], "System Alias": [[7, "module-olive.systems.system_alias"]], "Examples": [[8, "examples"], [9, "examples"], [9, "id1"]], "Custom Scripts": [[9, "custom-scripts"]], "user_script": [[9, "user-script"]], "script_dir": [[9, "script-dir"]], "Design": [[10, "design"]], "Pass": [[10, "pass"]], "System": [[10, "system"]], "Evaluator": [[10, "evaluator"], [39, "evaluator"]], "Search Strategy": [[10, "search-strategy"]], "Execution Order": [[10, "execution-order"]], "Search Algorithm": [[10, "search-algorithm"]], "How to add new optimization Pass": [[11, "how-to-add-new-optimization-pass"]], "1. Define a new class": [[11, "define-a-new-class"]], "2. Define configuration": [[11, "define-configuration"]], "Example": [[11, "example"], [37, "example"], [37, "id1"], [37, "id2"], [37, "id3"], [37, "id4"], [37, "id5"], [37, "id6"]], "3. Implement the run function": [[11, "implement-the-run-function"]], "Azure ML integration": [[12, "azure-ml-integration"]], "Azure Machine Learning client": [[12, "azure-machine-learning-client"]], "Olive Core": [[12, "olive-core"]], "Using AzureML registered model": [[12, "using-azureml-registered-model"]], "Using AzureML curated model": [[12, "using-azureml-curated-model"]], "Using model stored in AzureML datastore": [[12, "using-model-stored-in-azureml-datastore"]], "Using a model from an AzureML job output": [[12, "using-a-model-from-an-azureml-job-output"]], "Using data stored in AzureML datastore": [[12, "using-data-stored-in-azureml-datastore"]], "Using Azure ML compute as host or target": [[12, "using-azure-ml-compute-as-host-or-target"]], "Packaging the output models to AzureML workspace": [[12, "packaging-the-output-models-to-azureml-workspace"]], "Azure ML Models": [[12, "azure-ml-models"]], "Azure ML Data": [[12, "azure-ml-data"]], "Connect your own device to Azure ML as target or host by Azure Arc": [[12, "connect-your-own-device-to-azure-ml-as-target-or-host-by-azure-arc"]], "Azure ML Helper Scripts": [[12, "azure-ml-helper-scripts"]], "Command Line Tools": [[13, "command-line-tools"]], "Sub-commands": [[13, "Sub-commands"]], "run": [[13, "run"]], "Named Arguments": [[13, "named-arguments"], [13, "named-arguments_repeat1"], [13, "named-arguments_repeat2"], [13, "named-arguments_repeat3"], [13, "named-arguments_repeat4"], [13, "named-arguments_repeat5"], [13, "named-arguments_repeat6"]], "finetune": [[13, "finetune"]], "model options": [[13, "model-options"], [13, "model-options_repeat1"]], "dataset options": [[13, "dataset-options"]], "lora options": [[13, "lora-options"]], "remote options": [[13, "remote-options"]], "export-adapters": [[13, "export-adapters"]], "int4 quantization options": [[13, "int4-quantization-options"]], "configure-qualcomm-sdk": [[13, "configure-qualcomm-sdk"]], "manage-aml-compute": [[13, "manage-aml-compute"]], "tune-session-params": [[13, "tune-session-params"]], "dataset options, which mutually exclusive with huggingface dataset options": [[13, "dataset-options,-which-mutually-exclusive-with-huggingface-dataset-options"]], "huggingface dataset options, if dataset options are not provided, user should provide the following options to modify the default data config. Please refer to olive.data.container.TransformersTokenDummyDataContainer for more details.": [[13, "huggingface-dataset-options,-if-dataset-options-are-not-provided,-user-should-provide-the-following-options-to-modify-the-default-data-config.-please-refer-to-olive.data.container.transformerstokendummydatacontainer-for-more-details."]], "pass options": [[13, "pass-options"]], "cloud-cache": [[13, "cloud-cache"]], "Cloud Model Cache": [[14, "cloud-model-cache"]], "What is Cloud Model Cache": [[14, "what-is-cloud-model-cache"]], "How to use Cloud Model Cache": [[14, "how-to-use-cloud-model-cache"]], "Install Extra Dependencies": [[14, "install-extra-dependencies"], [30, "install-extra-dependencies"]], "Azure Blob Storage": [[14, "azure-blob-storage"]], "Cloud Cache Configuration": [[14, "cloud-cache-configuration"]], "Notes": [[14, "notes"]], "Model Conversions": [[15, "model-conversions"]], "Huggingface Integration": [[16, "huggingface-integration"]], "Introduction": [[16, "introduction"]], "Input Model": [[16, "input-model"], [38, "input-model"], [39, "input-model"]], "Huggingface Hub model": [[16, "huggingface-hub-model"]], "Local model": [[16, "local-model"]], "Azure ML model": [[16, "azure-ml-model"]], "Model config loading": [[16, "model-config-loading"]], "Huggingface datasets": [[16, "huggingface-datasets"]], "Huggingface metrics": [[16, "huggingface-metrics"]], "Huggingface login": [[16, "huggingface-login"]], "Local system, docker system and Python environment system": [[16, "local-system-docker-system-and-python-environment-system"]], "AzureML system": [[16, "azureml-system"]], "E2E example": [[16, "e2e-example"]], "LoRA Adapters": [[17, "lora-adapters"]], "Model Transformations and Optimizations": [[18, "model-transformations-and-optimizations"]], "Packaging Olive artifacts": [[19, "packaging-olive-artifacts"]], "What is Olive Packaging": [[19, "what-is-olive-packaging"]], "Zipfile": [[19, "zipfile"]], "CandidateModels": [[19, "candidatemodels"]], "Models rank JSON file": [[19, "models-rank-json-file"]], "AzureMLModels": [[19, "azuremlmodels"]], "AzureMLData": [[19, "azuremldata"]], "AzureMLDeployment": [[19, "azuremldeployment"]], "Dockerfile": [[19, "dockerfile"]], "How to package Olive artifacts": [[19, "how-to-package-olive-artifacts"]], "Packaged files": [[19, "packaged-files"]], "Inference config file": [[19, "inference-config-file"]], "Model configuration file": [[19, "model-configuration-file"]], "Metrics file": [[19, "metrics-file"]], "ONNX": [[20, "onnx"], [22, "onnx"], [26, "onnx"]], "Model Conversion": [[20, "model-conversion"], [23, "model-conversion"], [28, "model-conversion"]], "Example Configuration": [[20, "example-configuration"], [20, "id1"], [20, "id2"], [20, "id3"], [20, "id4"], [20, "id5"], [21, "example-configuration"], [22, "example-configuration"], [22, "id1"], [22, "id2"], [22, "id3"], [22, "id4"], [22, "id5"], [23, "example-configuration"], [23, "id1"], [24, "example-configuration"], [24, "id1"], [24, "id2"], [24, "id3"], [24, "id4"], [24, "id5"], [25, "example-configuration"], [25, "id1"], [25, "id2"], [26, "example-configuration"], [26, "id1"], [26, "id2"], [27, "example-configuration"], [27, "id1"], [27, "id2"], [28, "example-configuration"], [28, "id1"]], "Float16 Conversion": [[20, "float16-conversion"]], "Inputs/Outputs Float16 to Float32 Conversion": [[20, "inputs-outputs-float16-to-float32-conversion"]], "Mixed Precision Conversion": [[20, "mixed-precision-conversion"]], "Convert dynamic shape to fixed shape": [[20, "convert-dynamic-shape-to-fixed-shape"]], "PyTorch": [[21, "pytorch"], [24, "pytorch"], [27, "pytorch"]], "Model Optimizer": [[22, "model-optimizer"]], "ORT Transformers Optimization": [[22, "ort-transformers-optimization"]], "Append Pre/Post Processing Ops": [[22, "append-pre-post-processing-ops"]], "Insert Beam Search Op": [[22, "insert-beam-search-op"]], "ORT Performance Tuning": [[22, "ort-performance-tuning"]], "Extract Adapters": [[22, "extract-adapters"]], "OpenVINO": [[23, "openvino"]], "Prerequisites": [[23, "prerequisites"], [25, "prerequisites"], [28, "prerequisites"], [32, "prerequisites"], [34, "prerequisites"], [47, "prerequisites"], [47, "id2"]], "Option 1: install Olive with OpenVINO extras": [[23, "option-1-install-olive-with-openvino-extras"]], "Option 2: Install OpenVINO Runtime and OpenVINO Development Tools from Pypi": [[23, "option-2-install-openvino-runtime-and-openvino-development-tools-from-pypi"]], "Post Training Quantization (PTQ)": [[23, "post-training-quantization-ptq"], [28, "post-training-quantization-ptq"]], "Quantization Aware Training": [[24, "quantization-aware-training"]], "QNN": [[25, "qnn"]], "Download and unzip QNN SDK": [[25, "download-and-unzip-qnn-sdk"]], "Configure Olive QNN": [[25, "configure-olive-qnn"]], "Model Conversion/Quantization": [[25, "model-conversion-quantization"]], "Model Library Generation": [[25, "model-library-generation"]], "Context Binary Generation": [[25, "context-binary-generation"]], "Quantize with onnxruntime": [[26, "quantize-with-onnxruntime"]], "Quantize with Intel\u00ae Neural Compressor": [[26, "quantize-with-intel-neural-compressor"]], "Quantize with AMD Vitis AI Quantizer": [[26, "quantize-with-amd-vitis-ai-quantizer"]], "AutoGPTQ": [[27, "autogptq"]], "AutoAWQ": [[27, "autoawq"]], "SNPE": [[28, "snpe"]], "Download and unzip SNPE SDK": [[28, "download-and-unzip-snpe-sdk"]], "Configure Olive SNPE": [[28, "configure-olive-snpe"]], "Model Quantizations": [[29, "model-quantizations"]], "Remote Workflow": [[30, "remote-workflow"]], "Running Olive Workflow Remotely on Azure Machine Learning workspace compute": [[30, "running-olive-workflow-remotely-on-azure-machine-learning-workspace-compute"]], "Configure an AzureML system": [[30, "configure-an-azureml-system"]], "Run Olive Workflow": [[30, "run-olive-workflow"]], "Workflow outputs": [[30, "workflow-outputs"]], "Installation": [[31, "installation"]], "Install with pip": [[31, "install-with-pip"]], "Install from source": [[31, "install-from-source"]], "Editable install": [[31, "editable-install"]], "Optional Dependencies": [[31, "optional-dependencies"]], "BERT optimization with PTQ on CPU": [[32, "bert-optimization-with-ptq-on-cpu"]], "Pip requirements": [[32, "pip-requirements"], [34, "pip-requirements"]], "Run sample using config. The optimization techniques to run are specified in bert_ptq_cpu.json": [[32, "run-sample-using-config-the-optimization-techniques-to-run-are-specified-in-bert-ptq-cpu-json"]], "Optimize model automatically without selecting any optimization technique.": [[32, "optimize-model-automatically-without-selecting-any-optimization-technique"]], "Quickstart Examples": [[33, "quickstart-examples"]], "ResNet optimization with PTQ on CPU": [[34, "resnet-optimization-with-ptq-on-cpu"]], "Prepare data and model": [[34, "prepare-data-and-model"]], "Run sample using config": [[34, "run-sample-using-config"]], "Welcome to Olive\u2019s documentation!": [[35, "welcome-to-olive-s-documentation"]], "OVERVIEW": [[35, null]], "GET STARTED": [[35, null]], "EXAMPLES": [[35, null]], "FEATURES": [[35, null]], "EXTENDING OLIVE": [[35, null]], "TUTORIALS": [[35, null]], "API REFERENCE": [[35, null]], "Olive": [[36, "olive"]], "Contents": [[36, "contents"]], "Olive Options": [[37, "olive-options"]], "Workflow ID": [[37, "workflow-id"]], "Workflow Host": [[37, "workflow-host"]], "Azure ML Client": [[37, "azure-ml-client"]], "azureml_client with aml_config_path:": [[37, "azureml-client-with-aml-config-path"]], "aml_config.json:": [[37, "aml-config-json"]], "azureml_client:": [[37, "azureml-client"]], "azureml_client with azureml config fields:": [[37, "azureml-client-with-azureml-config-fields"]], "Input Model Information": [[37, "input-model-information"]], "Systems Information": [[37, "systems-information"]], "Evaluators Information": [[37, "evaluators-information"]], "Passes Information": [[37, "passes-information"]], "Pass Flows Information": [[37, "pass-flows-information"]], "Engine Information": [[37, "engine-information"]], "Quick Tour": [[38, "quick-tour"]], "Information needed to accelerate a model": [[38, "information-needed-to-accelerate-a-model"]], "Passes to apply": [[38, "passes-to-apply"]], "Olive Engine": [[38, "olive-engine"]], "Advanced User Tour": [[39, "advanced-user-tour"]], "Host and Target Systems": [[39, "host-and-target-systems"]], "Register Passes": [[39, "register-passes"]], "Run the engine": [[39, "run-the-engine"]], "Self-hosted Kubernetes cluster": [[40, "self-hosted-kubernetes-cluster"]], "Create Kubernetes cluster": [[40, "create-kubernetes-cluster"]], "Add local Kubernetes cluster to Azure Arc": [[40, "add-local-kubernetes-cluster-to-azure-arc"]], "Add AML extension to cluster": [[40, "add-aml-extension-to-cluster"]], "Link Azure Arc Kubernetes cluster to Azure Machine Learning": [[40, "link-azure-arc-kubernetes-cluster-to-azure-machine-learning"]], "Azure ML scripts": [[41, "azure-ml-scripts"]], "Scripts list": [[41, "scripts-list"]], "manage_compute_instance": [[41, "manage-compute-instance"]], "Usage": [[41, "usage"]], "How To Configure Auto Optimizer": [[42, "how-to-configure-auto-optimizer"]], "Auto Optimizer Configuration": [[42, "auto-optimizer-configuration"]], "How To Configure Data": [[43, "how-to-configure-data"]], "Supported Data Config Template": [[43, "supported-data-config-template"]], "Generic Data Config": [[43, "generic-data-config"]], "Title": [[43, "id2"]], "Configs with built-in component:": [[43, "configs-with-built-in-component"]], "Configs with customized component:": [[43, "configs-with-customized-component"]], "How To Configure Metric": [[44, "how-to-configure-metric"]], "Metric Types": [[44, "metric-types"]], "Accuracy Metric": [[44, "accuracy-metric"]], "Latency Metric": [[44, "latency-metric"]], "Throughput Metric": [[44, "throughput-metric"]], "Custom Metric": [[44, "custom-metric"]], "Multi Metrics configuration": [[44, "multi-metrics-configuration"]], "How To Set Model Path": [[45, "how-to-set-model-path"]], "Local Model Path": [[45, "local-model-path"]], "Local File": [[45, "local-file"]], "Local Folder": [[45, "local-folder"]], "String Name": [[45, "string-name"]], "Remote Model Path": [[45, "remote-model-path"]], "AzureML Model": [[45, "azureml-model"]], "AzureML Registry Model": [[45, "azureml-registry-model"]], "AzureML Datastore": [[45, "azureml-datastore"]], "AzureML Job Output": [[45, "azureml-job-output"]], "How To Configure Pass": [[46, "how-to-configure-pass"]], "How To Configure System": [[47, "how-to-configure-system"]], "Accelerator Configuration": [[47, "accelerator-configuration"]], "Local System": [[47, "local-system"]], "Configuration": [[47, null], [47, null], [47, null], [47, null], [47, null]], "Python Environment System": [[47, "python-environment-system"]], "Native Python Environment System": [[47, "native-python-environment-system"]], "Managed Python Environment System": [[47, "managed-python-environment-system"]], "Docker System": [[47, "docker-system"]], "Native Docker System": [[47, "native-docker-system"]], "Managed Docker System": [[47, "managed-docker-system"]], "AzureML System": [[47, "azureml-system"]], "Native AzureML System": [[47, "native-azureml-system"]], "AzureML Readymade Systems": [[47, "azureml-readymade-systems"]], "Managed AzureML System": [[47, "managed-azureml-system"]], "Isolated ORT System": [[47, "isolated-ort-system"]]}, "indexentries": {"engine (class in olive.engine)": [[0, "olive.engine.Engine"]], "searchstrategy (class in olive.strategy.search_strategy)": [[0, "olive.strategy.search_strategy.SearchStrategy"]], "cache_dir (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.cache_dir"]], "clean_cache (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.clean_cache"]], "clean_evaluation_cache (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.clean_evaluation_cache"]], "create_evaluator() (olive.evaluator.olive_evaluator.oliveevaluatorconfig method)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.create_evaluator"]], "evaluator (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.evaluator"]], "execution_order (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.execution_order"]], "host (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.host"]], "is_accuracy_drop_tolerance (olive.evaluator.olive_evaluator.oliveevaluatorconfig property)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.is_accuracy_drop_tolerance"]], "max_iter (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.max_iter"]], "max_time (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.max_time"]], "metrics (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.metrics"]], "output_model_num (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.output_model_num"]], "plot_pareto_frontier (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.plot_pareto_frontier"]], "register() (olive.engine.engine method)": [[0, "olive.engine.Engine.register"]], "run() (olive.engine.engine method)": [[0, "olive.engine.Engine.run"]], "script_dir (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.script_dir"]], "search_algorithm (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.search_algorithm"]], "search_algorithm_config (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.search_algorithm_config"]], "search_strategy (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.search_strategy"]], "stop_when_goals_met (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.stop_when_goals_met"]], "target (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.target"]], "type (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.type"]], "type_args (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.type_args"]], "user_script (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.user_script"]], "oliveevaluator (class in olive.evaluator.olive_evaluator)": [[1, "olive.evaluator.olive_evaluator.OliveEvaluator"]], "accuracy (olive.evaluator.metric.metrictype attribute)": [[2, "olive.evaluator.metric.MetricType.ACCURACY"]], "accuracy_score (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.ACCURACY_SCORE"]], "auroc (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.AUROC"]], "avg (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.AVG"]], "avg (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.AVG"]], "accuracysubtype (class in olive.evaluator.metric)": [[2, "olive.evaluator.metric.AccuracySubType"]], "custom (olive.evaluator.metric.metrictype attribute)": [[2, "olive.evaluator.metric.MetricType.CUSTOM"]], "f1_score (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.F1_SCORE"]], "latency (olive.evaluator.metric.metrictype attribute)": [[2, "olive.evaluator.metric.MetricType.LATENCY"]], "latencysubtype (class in olive.evaluator.metric)": [[2, "olive.evaluator.metric.LatencySubType"]], "max (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.MAX"]], "max (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.MAX"]], "min (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.MIN"]], "min (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.MIN"]], "metrictype (class in olive.evaluator.metric)": [[2, "olive.evaluator.metric.MetricType"]], "p50 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P50"]], "p50 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P50"]], "p75 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P75"]], "p75 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P75"]], "p90 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P90"]], "p90 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P90"]], "p95 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P95"]], "p95 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P95"]], "p99 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P99"]], "p99 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P99"]], "p999 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P999"]], "p999 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P999"]], "perplexity (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.PERPLEXITY"]], "precision (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.PRECISION"]], "recall (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.RECALL"]], "throughput (olive.evaluator.metric.metrictype attribute)": [[2, "olive.evaluator.metric.MetricType.THROUGHPUT"]], "throughputsubtype (class in olive.evaluator.metric)": [[2, "olive.evaluator.metric.ThroughputSubType"]], "backend (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.backend"]], "data_config (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.data_config"]], "get_inference_settings() (olive.evaluator.metric.metric method)": [[2, "olive.evaluator.metric.Metric.get_inference_settings"]], "get_run_kwargs() (olive.evaluator.metric.metric method)": [[2, "olive.evaluator.metric.Metric.get_run_kwargs"]], "get_sub_type_info() (olive.evaluator.metric.metric method)": [[2, "olive.evaluator.metric.Metric.get_sub_type_info"]], "has_regression_goal() (olive.evaluator.metric.metricgoal method)": [[2, "olive.evaluator.metric.MetricGoal.has_regression_goal"]], "name (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.name"]], "sub_types (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.sub_types"]], "type (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.type"]], "type (olive.evaluator.metric.metricgoal attribute)": [[2, "olive.evaluator.metric.MetricGoal.type"]], "user_config (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.user_config"]], "value (olive.evaluator.metric.metricgoal attribute)": [[2, "olive.evaluator.metric.MetricGoal.value"]], "compositemodelhandler (class in olive.model)": [[3, "olive.model.CompositeModelHandler"]], "distributedhfmodelhandler (class in olive.model)": [[3, "olive.model.DistributedHfModelHandler"]], "distributedonnxmodelhandler (class in olive.model)": [[3, "olive.model.DistributedOnnxModelHandler"]], "hfmodelhandler (class in olive.model)": [[3, "olive.model.HfModelHandler"]], "modelconfig (class in olive.model)": [[3, "olive.model.ModelConfig"]], "onnxmodelhandler (class in olive.model)": [[3, "olive.model.ONNXModelHandler"]], "openvinomodelhandler (class in olive.model)": [[3, "olive.model.OpenVINOModelHandler"]], "pytorchmodelhandler (class in olive.model)": [[3, "olive.model.PyTorchModelHandler"]], "snpemodelhandler (class in olive.model)": [[3, "olive.model.SNPEModelHandler"]], "activationsymmetric": [[4, "cmdoption-arg-164"], [4, "cmdoption-arg-62"], [4, "cmdoption-arg-89"], [4, "cmdoption-arg-ActivationSymmetric"]], "addqdqpairtoweight": [[4, "cmdoption-arg-AddQDQPairToWeight"]], "enablesubgraph": [[4, "cmdoption-arg-64"], [4, "cmdoption-arg-91"], [4, "cmdoption-arg-EnableSubgraph"]], "forcequantizenoinputcheck": [[4, "cmdoption-arg-65"], [4, "cmdoption-arg-92"], [4, "cmdoption-arg-ForceQuantizeNoInputCheck"]], "matmulconstbonly": [[4, "cmdoption-arg-66"], [4, "cmdoption-arg-93"], [4, "cmdoption-arg-MatMulConstBOnly"]], "weightsymmetric": [[4, "cmdoption-arg-165"], [4, "cmdoption-arg-63"], [4, "cmdoption-arg-90"], [4, "cmdoption-arg-WeightSymmetric"]], "a_asym": [[4, "cmdoption-arg-a_asym"]], "a_bits": [[4, "cmdoption-arg-a_bits"]], "a_clip_ratio": [[4, "cmdoption-arg-a_clip_ratio"]], "a_groupsize": [[4, "cmdoption-arg-a_groupsize"]], "activation_type": [[4, "cmdoption-arg-163"], [4, "cmdoption-arg-85"], [4, "cmdoption-arg-activation_type"]], "all_tensors_to_one_file": [[4, "cmdoption-arg-101"], [4, "cmdoption-arg-108"], [4, "cmdoption-arg-12"], [4, "cmdoption-arg-128"], [4, "cmdoption-arg-149"], [4, "cmdoption-arg-168"], [4, "cmdoption-arg-174"], [4, "cmdoption-arg-180"], [4, "cmdoption-arg-185"], [4, "cmdoption-arg-2"], [4, "cmdoption-arg-21"], [4, "cmdoption-arg-224"], [4, "cmdoption-arg-259"], [4, "cmdoption-arg-26"], [4, "cmdoption-arg-32"], [4, "cmdoption-arg-37"], [4, "cmdoption-arg-42"], [4, "cmdoption-arg-47"], [4, "cmdoption-arg-69"], [4, "cmdoption-arg-7"], [4, "cmdoption-arg-96"], [4, "cmdoption-arg-all_tensors_to_one_file"]], "allow_tf32": [[4, "cmdoption-arg-195"], [4, "cmdoption-arg-205"], [4, "cmdoption-arg-allow_tf32"]], "append_first_op_types_to_quantize_list": [[4, "cmdoption-arg-54"], [4, "cmdoption-arg-76"], [4, "cmdoption-arg-append_first_op_types_to_quantize_list"]], "approach": [[4, "cmdoption-arg-112"], [4, "cmdoption-arg-132"], [4, "cmdoption-arg-approach"]], "atol": [[4, "cmdoption-arg-atol"]], "backend": [[4, "cmdoption-arg-114"], [4, "cmdoption-arg-134"], [4, "cmdoption-arg-229"], [4, "cmdoption-arg-backend"]], "binary_file": [[4, "cmdoption-arg-binary_file"]], "bits": [[4, "cmdoption-arg-bits"]], "blocksize": [[4, "cmdoption-arg-blocksize"]], "calibrate_method": [[4, "cmdoption-arg-161"], [4, "cmdoption-arg-83"], [4, "cmdoption-arg-calibrate_method"]], "calibration_batch_size": [[4, "cmdoption-arg-237"], [4, "cmdoption-arg-calibration_batch_size"]], "calibration_data_config": [[4, "cmdoption-arg-235"], [4, "cmdoption-arg-calibration_data_config"]], "calibration_nsamples": [[4, "cmdoption-arg-236"], [4, "cmdoption-arg-calibration_nsamples"]], "calibration_sampling_size": [[4, "cmdoption-arg-147"], [4, "cmdoption-arg-calibration_sampling_size"]], "checkpoint_path": [[4, "cmdoption-arg-checkpoint_path"]], "command line option": [[4, "cmdoption-arg-0"], [4, "cmdoption-arg-1"], [4, "cmdoption-arg-10"], [4, "cmdoption-arg-100"], [4, "cmdoption-arg-101"], [4, "cmdoption-arg-102"], [4, "cmdoption-arg-103"], [4, "cmdoption-arg-104"], [4, "cmdoption-arg-105"], [4, "cmdoption-arg-106"], [4, "cmdoption-arg-107"], [4, "cmdoption-arg-108"], [4, "cmdoption-arg-109"], [4, "cmdoption-arg-11"], [4, "cmdoption-arg-110"], [4, "cmdoption-arg-111"], [4, "cmdoption-arg-112"], [4, "cmdoption-arg-113"], [4, "cmdoption-arg-114"], [4, "cmdoption-arg-115"], [4, "cmdoption-arg-116"], [4, "cmdoption-arg-117"], [4, "cmdoption-arg-118"], [4, "cmdoption-arg-119"], [4, "cmdoption-arg-12"], [4, "cmdoption-arg-120"], [4, "cmdoption-arg-121"], [4, "cmdoption-arg-122"], [4, "cmdoption-arg-123"], [4, "cmdoption-arg-124"], [4, "cmdoption-arg-125"], [4, "cmdoption-arg-126"], [4, "cmdoption-arg-127"], [4, "cmdoption-arg-128"], [4, "cmdoption-arg-129"], [4, "cmdoption-arg-13"], [4, "cmdoption-arg-130"], [4, "cmdoption-arg-131"], [4, "cmdoption-arg-132"], [4, "cmdoption-arg-133"], [4, "cmdoption-arg-134"], [4, "cmdoption-arg-135"], [4, "cmdoption-arg-136"], [4, "cmdoption-arg-137"], [4, "cmdoption-arg-138"], [4, "cmdoption-arg-139"], [4, "cmdoption-arg-14"], [4, "cmdoption-arg-140"], [4, "cmdoption-arg-141"], [4, "cmdoption-arg-142"], [4, "cmdoption-arg-143"], [4, "cmdoption-arg-144"], [4, "cmdoption-arg-145"], [4, "cmdoption-arg-146"], [4, "cmdoption-arg-147"], [4, "cmdoption-arg-148"], [4, "cmdoption-arg-149"], [4, "cmdoption-arg-15"], [4, "cmdoption-arg-150"], [4, "cmdoption-arg-151"], [4, "cmdoption-arg-152"], [4, "cmdoption-arg-153"], [4, "cmdoption-arg-154"], [4, "cmdoption-arg-155"], [4, "cmdoption-arg-156"], [4, "cmdoption-arg-157"], [4, "cmdoption-arg-158"], [4, "cmdoption-arg-159"], [4, "cmdoption-arg-16"], [4, "cmdoption-arg-160"], [4, "cmdoption-arg-161"], [4, "cmdoption-arg-162"], [4, "cmdoption-arg-163"], [4, "cmdoption-arg-164"], [4, "cmdoption-arg-165"], [4, "cmdoption-arg-166"], [4, "cmdoption-arg-167"], [4, "cmdoption-arg-168"], [4, "cmdoption-arg-169"], [4, "cmdoption-arg-17"], [4, "cmdoption-arg-170"], [4, "cmdoption-arg-171"], [4, "cmdoption-arg-172"], [4, "cmdoption-arg-173"], [4, "cmdoption-arg-174"], [4, "cmdoption-arg-175"], [4, "cmdoption-arg-176"], [4, "cmdoption-arg-177"], [4, "cmdoption-arg-178"], [4, "cmdoption-arg-179"], [4, "cmdoption-arg-18"], [4, "cmdoption-arg-180"], [4, "cmdoption-arg-181"], [4, "cmdoption-arg-182"], [4, "cmdoption-arg-183"], [4, "cmdoption-arg-184"], [4, "cmdoption-arg-185"], [4, "cmdoption-arg-186"], [4, "cmdoption-arg-187"], [4, "cmdoption-arg-188"], [4, "cmdoption-arg-189"], [4, "cmdoption-arg-19"], [4, "cmdoption-arg-190"], [4, "cmdoption-arg-191"], [4, "cmdoption-arg-192"], [4, "cmdoption-arg-193"], [4, "cmdoption-arg-194"], [4, "cmdoption-arg-195"], [4, "cmdoption-arg-196"], [4, "cmdoption-arg-197"], [4, "cmdoption-arg-198"], [4, "cmdoption-arg-199"], [4, "cmdoption-arg-2"], [4, "cmdoption-arg-20"], [4, "cmdoption-arg-200"], [4, "cmdoption-arg-201"], [4, "cmdoption-arg-202"], [4, "cmdoption-arg-203"], [4, "cmdoption-arg-204"], [4, "cmdoption-arg-205"], [4, "cmdoption-arg-206"], [4, "cmdoption-arg-207"], [4, "cmdoption-arg-208"], [4, "cmdoption-arg-209"], [4, "cmdoption-arg-21"], [4, "cmdoption-arg-210"], [4, "cmdoption-arg-211"], [4, "cmdoption-arg-212"], [4, "cmdoption-arg-213"], [4, "cmdoption-arg-214"], [4, "cmdoption-arg-215"], [4, "cmdoption-arg-216"], [4, "cmdoption-arg-217"], [4, "cmdoption-arg-218"], [4, "cmdoption-arg-219"], [4, "cmdoption-arg-22"], [4, "cmdoption-arg-220"], [4, "cmdoption-arg-221"], [4, "cmdoption-arg-222"], [4, "cmdoption-arg-223"], [4, "cmdoption-arg-224"], [4, "cmdoption-arg-225"], [4, "cmdoption-arg-226"], [4, "cmdoption-arg-227"], [4, "cmdoption-arg-228"], [4, "cmdoption-arg-229"], [4, "cmdoption-arg-23"], [4, "cmdoption-arg-230"], [4, "cmdoption-arg-231"], [4, "cmdoption-arg-232"], [4, "cmdoption-arg-233"], [4, "cmdoption-arg-234"], [4, "cmdoption-arg-235"], [4, "cmdoption-arg-236"], [4, "cmdoption-arg-237"], [4, "cmdoption-arg-238"], [4, "cmdoption-arg-239"], [4, "cmdoption-arg-24"], [4, "cmdoption-arg-240"], [4, "cmdoption-arg-241"], [4, "cmdoption-arg-242"], [4, "cmdoption-arg-243"], [4, "cmdoption-arg-244"], [4, "cmdoption-arg-245"], [4, "cmdoption-arg-246"], [4, "cmdoption-arg-247"], [4, "cmdoption-arg-248"], [4, "cmdoption-arg-249"], [4, "cmdoption-arg-25"], [4, "cmdoption-arg-250"], [4, "cmdoption-arg-251"], [4, "cmdoption-arg-252"], [4, "cmdoption-arg-253"], [4, "cmdoption-arg-254"], [4, "cmdoption-arg-255"], [4, "cmdoption-arg-256"], [4, "cmdoption-arg-257"], [4, "cmdoption-arg-258"], [4, "cmdoption-arg-259"], [4, "cmdoption-arg-26"], [4, "cmdoption-arg-260"], [4, "cmdoption-arg-261"], [4, "cmdoption-arg-262"], [4, "cmdoption-arg-263"], [4, "cmdoption-arg-27"], [4, "cmdoption-arg-28"], [4, "cmdoption-arg-29"], [4, "cmdoption-arg-3"], [4, "cmdoption-arg-30"], [4, "cmdoption-arg-31"], [4, "cmdoption-arg-32"], [4, "cmdoption-arg-33"], [4, "cmdoption-arg-34"], [4, "cmdoption-arg-35"], [4, "cmdoption-arg-36"], [4, "cmdoption-arg-37"], [4, "cmdoption-arg-38"], [4, "cmdoption-arg-39"], [4, "cmdoption-arg-4"], [4, "cmdoption-arg-40"], [4, "cmdoption-arg-41"], [4, "cmdoption-arg-42"], [4, "cmdoption-arg-43"], [4, "cmdoption-arg-44"], [4, "cmdoption-arg-45"], [4, "cmdoption-arg-46"], [4, "cmdoption-arg-47"], [4, "cmdoption-arg-48"], [4, "cmdoption-arg-49"], [4, "cmdoption-arg-5"], [4, "cmdoption-arg-50"], [4, "cmdoption-arg-51"], [4, "cmdoption-arg-52"], [4, "cmdoption-arg-53"], [4, "cmdoption-arg-54"], [4, "cmdoption-arg-55"], [4, "cmdoption-arg-56"], [4, "cmdoption-arg-57"], [4, "cmdoption-arg-58"], [4, "cmdoption-arg-59"], [4, "cmdoption-arg-6"], [4, "cmdoption-arg-60"], [4, "cmdoption-arg-61"], [4, "cmdoption-arg-62"], [4, "cmdoption-arg-63"], [4, "cmdoption-arg-64"], [4, "cmdoption-arg-65"], [4, "cmdoption-arg-66"], [4, "cmdoption-arg-67"], [4, "cmdoption-arg-68"], [4, "cmdoption-arg-69"], [4, "cmdoption-arg-7"], [4, "cmdoption-arg-70"], [4, "cmdoption-arg-71"], [4, "cmdoption-arg-72"], [4, "cmdoption-arg-73"], [4, "cmdoption-arg-74"], [4, "cmdoption-arg-75"], [4, "cmdoption-arg-76"], [4, "cmdoption-arg-77"], [4, "cmdoption-arg-78"], [4, "cmdoption-arg-79"], [4, "cmdoption-arg-8"], [4, "cmdoption-arg-80"], [4, "cmdoption-arg-81"], [4, "cmdoption-arg-82"], [4, "cmdoption-arg-83"], [4, "cmdoption-arg-84"], [4, "cmdoption-arg-85"], [4, "cmdoption-arg-86"], [4, "cmdoption-arg-87"], [4, "cmdoption-arg-88"], [4, "cmdoption-arg-89"], [4, "cmdoption-arg-9"], [4, "cmdoption-arg-90"], [4, "cmdoption-arg-91"], [4, "cmdoption-arg-92"], [4, "cmdoption-arg-93"], [4, "cmdoption-arg-94"], [4, "cmdoption-arg-95"], [4, "cmdoption-arg-96"], [4, "cmdoption-arg-97"], [4, "cmdoption-arg-98"], [4, "cmdoption-arg-99"], [4, "cmdoption-arg-ActivationSymmetric"], [4, "cmdoption-arg-AddQDQPairToWeight"], [4, "cmdoption-arg-EnableSubgraph"], [4, "cmdoption-arg-ForceQuantizeNoInputCheck"], [4, "cmdoption-arg-MatMulConstBOnly"], [4, "cmdoption-arg-WeightSymmetric"], [4, "cmdoption-arg-a_asym"], [4, "cmdoption-arg-a_bits"], [4, "cmdoption-arg-a_clip_ratio"], [4, "cmdoption-arg-a_groupsize"], [4, "cmdoption-arg-activation_type"], [4, "cmdoption-arg-all_tensors_to_one_file"], [4, "cmdoption-arg-allow_tf32"], [4, "cmdoption-arg-append_first_op_types_to_quantize_list"], [4, "cmdoption-arg-approach"], [4, "cmdoption-arg-atol"], [4, "cmdoption-arg-backend"], [4, "cmdoption-arg-binary_file"], [4, "cmdoption-arg-bits"], [4, "cmdoption-arg-blocksize"], [4, "cmdoption-arg-calibrate_method"], [4, "cmdoption-arg-calibration_batch_size"], [4, "cmdoption-arg-calibration_data_config"], [4, "cmdoption-arg-calibration_nsamples"], [4, "cmdoption-arg-calibration_sampling_size"], [4, "cmdoption-arg-checkpoint_path"], [4, "cmdoption-arg-components"], [4, "cmdoption-arg-compress_to_fp16"], [4, "cmdoption-arg-compute_dtype"], [4, "cmdoption-arg-convert_attribute"], [4, "cmdoption-arg-cpu_cores"], [4, "cmdoption-arg-damp_percent"], [4, "cmdoption-arg-data_config"], [4, "cmdoption-arg-desc_act"], [4, "cmdoption-arg-device"], [4, "cmdoption-arg-dim_param"], [4, "cmdoption-arg-dim_value"], [4, "cmdoption-arg-do_validate"], [4, "cmdoption-arg-domain"], [4, "cmdoption-arg-double_quant"], [4, "cmdoption-arg-duo_scaling"], [4, "cmdoption-arg-dynamic_lora_r"], [4, "cmdoption-arg-element_wise_binary_ops"], [4, "cmdoption-arg-enable_cuda_graph"], [4, "cmdoption-arg-enable_dpu"], [4, "cmdoption-arg-enable_htp"], [4, "cmdoption-arg-enable_profiling"], [4, "cmdoption-arg-eval_data_config"], [4, "cmdoption-arg-example_input_func"], [4, "cmdoption-arg-exclude_embeds"], [4, "cmdoption-arg-exclude_lm_head"], [4, "cmdoption-arg-excluded_precisions"], [4, "cmdoption-arg-execution_mode_list"], [4, "cmdoption-arg-export_compatible"], [4, "cmdoption-arg-external_data_name"], [4, "cmdoption-arg-extra.Sigmoid.nnapi"], [4, "cmdoption-arg-extra_args"], [4, "cmdoption-arg-extra_configs"], [4, "cmdoption-arg-extra_options"], [4, "cmdoption-arg-extra_session_config"], [4, "cmdoption-arg-final_orientation"], [4, "cmdoption-arg-float16"], [4, "cmdoption-arg-force_evaluate_other_eps"], [4, "cmdoption-arg-force_fp16_inputs"], [4, "cmdoption-arg-force_fp32_nodes"], [4, "cmdoption-arg-force_fp32_ops"], [4, "cmdoption-arg-fp16"], [4, "cmdoption-arg-fuse_layernorm"], [4, "cmdoption-arg-gptq_damping"], [4, "cmdoption-arg-gptq_opt_scales"], [4, "cmdoption-arg-gpus"], [4, "cmdoption-arg-group_size"], [4, "cmdoption-arg-hidden_size"], [4, "cmdoption-arg-htp_socs"], [4, "cmdoption-arg-ignored_scope"], [4, "cmdoption-arg-ignored_scope_type"], [4, "cmdoption-arg-input"], [4, "cmdoption-arg-input_dim"], [4, "cmdoption-arg-input_int32"], [4, "cmdoption-arg-input_layouts"], [4, "cmdoption-arg-input_model_dtype"], [4, "cmdoption-arg-input_name"], [4, "cmdoption-arg-input_names"], [4, "cmdoption-arg-input_nodes"], [4, "cmdoption-arg-input_shape"], [4, "cmdoption-arg-input_shapes"], [4, "cmdoption-arg-input_types"], [4, "cmdoption-arg-inputs_to_make_channel_last"], [4, "cmdoption-arg-inside_layer_modules"], [4, "cmdoption-arg-int4_accuracy_level"], [4, "cmdoption-arg-int4_block_size"], [4, "cmdoption-arg-inter_thread_num_list"], [4, "cmdoption-arg-intra_thread_num_list"], [4, "cmdoption-arg-io_bind"], [4, "cmdoption-arg-k_bits"], [4, "cmdoption-arg-k_clip_ratio"], [4, "cmdoption-arg-k_groupsize"], [4, "cmdoption-arg-keep_io_types"], [4, "cmdoption-arg-layer_name_filter"], [4, "cmdoption-arg-layers_block_name"], [4, "cmdoption-arg-lib_name"], [4, "cmdoption-arg-lib_targets"], [4, "cmdoption-arg-loftq_iter"], [4, "cmdoption-arg-logger"], [4, "cmdoption-arg-lora_alpha"], [4, "cmdoption-arg-lora_dropout"], [4, "cmdoption-arg-lora_r"], [4, "cmdoption-arg-make_inputs"], [4, "cmdoption-arg-max_finite_val"], [4, "cmdoption-arg-max_layer"], [4, "cmdoption-arg-merge_adapter_weights"], [4, "cmdoption-arg-metadata_only"], [4, "cmdoption-arg-metric"], [4, "cmdoption-arg-min_layer"], [4, "cmdoption-arg-min_positive_val"], [4, "cmdoption-arg-model_type"], [4, "cmdoption-arg-modules_to_fuse"], [4, "cmdoption-arg-modules_to_not_convert"], [4, "cmdoption-arg-modules_to_save"], [4, "cmdoption-arg-name_pattern"], [4, "cmdoption-arg-need_layer_fusing"], [4, "cmdoption-arg-no_repeat_ngram_size"], [4, "cmdoption-arg-node_block_list"], [4, "cmdoption-arg-nodes_to_exclude"], [4, "cmdoption-arg-nodes_to_quantize"], [4, "cmdoption-arg-nsamples"], [4, "cmdoption-arg-num_epochs"], [4, "cmdoption-arg-num_heads"], [4, "cmdoption-arg-num_key_value_heads"], [4, "cmdoption-arg-num_steps"], [4, "cmdoption-arg-only_onnxruntime"], [4, "cmdoption-arg-op_block_list"], [4, "cmdoption-arg-op_type_dict"], [4, "cmdoption-arg-op_types_to_quantize"], [4, "cmdoption-arg-opt_level"], [4, "cmdoption-arg-opt_level_list"], [4, "cmdoption-arg-optimization_options"], [4, "cmdoption-arg-optimize_model"], [4, "cmdoption-arg-optional_inputs"], [4, "cmdoption-arg-out_node"], [4, "cmdoption-arg-output_model"], [4, "cmdoption-arg-output_names"], [4, "cmdoption-arg-output_nodes"], [4, "cmdoption-arg-outputs_to_make_channel_last"], [4, "cmdoption-arg-outside_layer_modules"], [4, "cmdoption-arg-overrides_config"], [4, "cmdoption-arg-pack_inputs"], [4, "cmdoption-arg-pack_model_for_onnx_conversion"], [4, "cmdoption-arg-parallel_jobs"], [4, "cmdoption-arg-past_key_value_name"], [4, "cmdoption-arg-per_channel"], [4, "cmdoption-arg-percdamp"], [4, "cmdoption-arg-post"], [4, "cmdoption-arg-pre"], [4, "cmdoption-arg-precision"], [4, "cmdoption-arg-prepare_qnn_config"], [4, "cmdoption-arg-preset"], [4, "cmdoption-arg-providers_list"], [4, "cmdoption-arg-ptl_data_module"], [4, "cmdoption-arg-ptl_module"], [4, "cmdoption-arg-q_group_size"], [4, "cmdoption-arg-qconfig_func"], [4, "cmdoption-arg-qnn_extra_options"], [4, "cmdoption-arg-quant_format"], [4, "cmdoption-arg-quant_level"], [4, "cmdoption-arg-quant_mode"], [4, "cmdoption-arg-quant_preprocess"], [4, "cmdoption-arg-quant_type"], [4, "cmdoption-arg-recipes"], [4, "cmdoption-arg-reduce_range"], [4, "cmdoption-arg-rotate"], [4, "cmdoption-arg-rotation_seed"], [4, "cmdoption-arg-round_interval"], [4, "cmdoption-arg-save_as_external_data"], [4, "cmdoption-arg-save_format"], [4, "cmdoption-arg-save_metadata_for_token_generation"], [4, "cmdoption-arg-script_dir"], [4, "cmdoption-arg-search"], [4, "cmdoption-arg-seed"], [4, "cmdoption-arg-size_threshold"], [4, "cmdoption-arg-sparsity"], [4, "cmdoption-arg-static_groups"], [4, "cmdoption-arg-strict"], [4, "cmdoption-arg-sym"], [4, "cmdoption-arg-target_device"], [4, "cmdoption-arg-target_modules"], [4, "cmdoption-arg-target_opset"], [4, "cmdoption-arg-tool_command"], [4, "cmdoption-arg-tool_command_args"], [4, "cmdoption-arg-torch_dtype"], [4, "cmdoption-arg-train_data_config"], [4, "cmdoption-arg-training_args"], [4, "cmdoption-arg-training_loop_func"], [4, "cmdoption-arg-trt_fp16_enable"], [4, "cmdoption-arg-true_sequential"], [4, "cmdoption-arg-tuning_criterion"], [4, "cmdoption-arg-use_dynamo_exporter"], [4, "cmdoption-arg-use_enhanced_quantizer"], [4, "cmdoption-arg-use_external_data_format"], [4, "cmdoption-arg-use_forced_decoder_ids"], [4, "cmdoption-arg-use_gpu"], [4, "cmdoption-arg-use_gqa"], [4, "cmdoption-arg-use_logits_processor"], [4, "cmdoption-arg-use_prefix_vocab_mask"], [4, "cmdoption-arg-use_symbolic_shape_infer"], [4, "cmdoption-arg-use_temperature"], [4, "cmdoption-arg-use_vocab_mask"], [4, "cmdoption-arg-user_script"], [4, "cmdoption-arg-v_bits"], [4, "cmdoption-arg-v_clip_ratio"], [4, "cmdoption-arg-v_groupsize"], [4, "cmdoption-arg-val_data_config"], [4, "cmdoption-arg-version"], [4, "cmdoption-arg-w_asym"], [4, "cmdoption-arg-w_bit"], [4, "cmdoption-arg-w_bits"], [4, "cmdoption-arg-w_gptq"], [4, "cmdoption-arg-w_groupsize"], [4, "cmdoption-arg-w_rtn"], [4, "cmdoption-arg-weight_only_config"], [4, "cmdoption-arg-weight_type"], [4, "cmdoption-arg-workspace"], [4, "cmdoption-arg-zero_point"], [5, "cmdoption-arg-0"], [5, "cmdoption-arg-1"], [5, "cmdoption-arg-2"], [5, "cmdoption-arg-3"], [5, "cmdoption-arg-4"], [5, "cmdoption-arg-azureml_client"], [5, "cmdoption-arg-datastore_name"], [5, "cmdoption-arg-datastore_url"], [5, "cmdoption-arg-job_name"], [5, "cmdoption-arg-name"], [5, "cmdoption-arg-output_name"], [5, "cmdoption-arg-path"], [5, "cmdoption-arg-relative_path"], [5, "cmdoption-arg-version"], [6, "cmdoption-arg-0"], [6, "cmdoption-arg-1"], [6, "cmdoption-arg-group"], [6, "cmdoption-arg-multivariate"], [6, "cmdoption-arg-num_samples"], [6, "cmdoption-arg-seed"], [6, "cmdoption-arg-with_replacement"]], "components": [[4, "cmdoption-arg-components"]], "compress_to_fp16": [[4, "cmdoption-arg-compress_to_fp16"]], "compute_dtype": [[4, "cmdoption-arg-199"], [4, "cmdoption-arg-compute_dtype"]], "convert_attribute": [[4, "cmdoption-arg-10"], [4, "cmdoption-arg-104"], [4, "cmdoption-arg-111"], [4, "cmdoption-arg-131"], [4, "cmdoption-arg-15"], [4, "cmdoption-arg-152"], [4, "cmdoption-arg-171"], [4, "cmdoption-arg-177"], [4, "cmdoption-arg-183"], [4, "cmdoption-arg-188"], [4, "cmdoption-arg-227"], [4, "cmdoption-arg-24"], [4, "cmdoption-arg-262"], [4, "cmdoption-arg-29"], [4, "cmdoption-arg-35"], [4, "cmdoption-arg-40"], [4, "cmdoption-arg-45"], [4, "cmdoption-arg-5"], [4, "cmdoption-arg-50"], [4, "cmdoption-arg-72"], [4, "cmdoption-arg-99"], [4, "cmdoption-arg-convert_attribute"]], "cpu_cores": [[4, "cmdoption-arg-cpu_cores"]], "create_training_args() (olive.passes.pytorch.lora.hftrainingarguments method)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.create_training_args"]], "damp_percent": [[4, "cmdoption-arg-damp_percent"]], "data_config": [[4, "cmdoption-arg-125"], [4, "cmdoption-arg-145"], [4, "cmdoption-arg-154"], [4, "cmdoption-arg-216"], [4, "cmdoption-arg-219"], [4, "cmdoption-arg-232"], [4, "cmdoption-arg-242"], [4, "cmdoption-arg-246"], [4, "cmdoption-arg-251"], [4, "cmdoption-arg-60"], [4, "cmdoption-arg-82"], [4, "cmdoption-arg-data_config"]], "desc_act": [[4, "cmdoption-arg-desc_act"]], "device": [[4, "cmdoption-arg-105"], [4, "cmdoption-arg-113"], [4, "cmdoption-arg-133"], [4, "cmdoption-arg-18"], [4, "cmdoption-arg-231"], [4, "cmdoption-arg-256"], [4, "cmdoption-arg-device"]], "dim_param": [[4, "cmdoption-arg-dim_param"]], "dim_value": [[4, "cmdoption-arg-dim_value"]], "do_validate": [[4, "cmdoption-arg-do_validate"]], "domain": [[4, "cmdoption-arg-115"], [4, "cmdoption-arg-135"], [4, "cmdoption-arg-domain"]], "double_quant": [[4, "cmdoption-arg-double_quant"]], "duo_scaling": [[4, "cmdoption-arg-duo_scaling"]], "dynamic_lora_r": [[4, "cmdoption-arg-dynamic_lora_r"]], "element_wise_binary_ops": [[4, "cmdoption-arg-element_wise_binary_ops"]], "enable_cuda_graph": [[4, "cmdoption-arg-263"], [4, "cmdoption-arg-enable_cuda_graph"]], "enable_dpu": [[4, "cmdoption-arg-enable_dpu"]], "enable_htp": [[4, "cmdoption-arg-enable_htp"]], "enable_profiling": [[4, "cmdoption-arg-enable_profiling"]], "eval_data_config": [[4, "cmdoption-arg-197"], [4, "cmdoption-arg-207"], [4, "cmdoption-arg-eval_data_config"]], "evaluation_strategy (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.evaluation_strategy"]], "example_input_func": [[4, "cmdoption-arg-example_input_func"]], "exclude_embeds": [[4, "cmdoption-arg-exclude_embeds"]], "exclude_lm_head": [[4, "cmdoption-arg-exclude_lm_head"]], "excluded_precisions": [[4, "cmdoption-arg-120"], [4, "cmdoption-arg-140"], [4, "cmdoption-arg-excluded_precisions"]], "execution_mode_list": [[4, "cmdoption-arg-execution_mode_list"]], "export_compatible": [[4, "cmdoption-arg-export_compatible"]], "external_data_name": [[4, "cmdoption-arg-102"], [4, "cmdoption-arg-109"], [4, "cmdoption-arg-129"], [4, "cmdoption-arg-13"], [4, "cmdoption-arg-150"], [4, "cmdoption-arg-169"], [4, "cmdoption-arg-175"], [4, "cmdoption-arg-181"], [4, "cmdoption-arg-186"], [4, "cmdoption-arg-22"], [4, "cmdoption-arg-225"], [4, "cmdoption-arg-260"], [4, "cmdoption-arg-27"], [4, "cmdoption-arg-3"], [4, "cmdoption-arg-33"], [4, "cmdoption-arg-38"], [4, "cmdoption-arg-43"], [4, "cmdoption-arg-48"], [4, "cmdoption-arg-70"], [4, "cmdoption-arg-8"], [4, "cmdoption-arg-97"], [4, "cmdoption-arg-external_data_name"]], "extra.sigmoid.nnapi": [[4, "cmdoption-arg-61"], [4, "cmdoption-arg-88"], [4, "cmdoption-arg-extra.Sigmoid.nnapi"]], "extra_args": [[4, "cmdoption-arg-220"], [4, "cmdoption-arg-228"], [4, "cmdoption-arg-230"], [4, "cmdoption-arg-257"], [4, "cmdoption-arg-extra_args"]], "extra_args (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.extra_args"]], "extra_configs": [[4, "cmdoption-arg-218"], [4, "cmdoption-arg-extra_configs"]], "extra_options": [[4, "cmdoption-arg-166"], [4, "cmdoption-arg-67"], [4, "cmdoption-arg-94"], [4, "cmdoption-arg-extra_options"]], "extra_session_config": [[4, "cmdoption-arg-extra_session_config"]], "final_orientation": [[4, "cmdoption-arg-final_orientation"]], "float16": [[4, "cmdoption-arg-250"], [4, "cmdoption-arg-float16"]], "force_evaluate_other_eps": [[4, "cmdoption-arg-force_evaluate_other_eps"]], "force_fp16_inputs": [[4, "cmdoption-arg-force_fp16_inputs"]], "force_fp32_nodes": [[4, "cmdoption-arg-force_fp32_nodes"]], "force_fp32_ops": [[4, "cmdoption-arg-force_fp32_ops"]], "fp16": [[4, "cmdoption-arg-255"], [4, "cmdoption-arg-fp16"]], "fuse_layernorm": [[4, "cmdoption-arg-fuse_layernorm"]], "gptq_damping": [[4, "cmdoption-arg-gptq_damping"]], "gptq_opt_scales": [[4, "cmdoption-arg-gptq_opt_scales"]], "gpus": [[4, "cmdoption-arg-gpus"]], "gradient_checkpointing (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.gradient_checkpointing"]], "group_size": [[4, "cmdoption-arg-group_size"]], "hidden_size": [[4, "cmdoption-arg-hidden_size"]], "htp_socs": [[4, "cmdoption-arg-htp_socs"]], "ignored_scope": [[4, "cmdoption-arg-ignored_scope"]], "ignored_scope_type": [[4, "cmdoption-arg-ignored_scope_type"]], "input": [[4, "cmdoption-arg-input"]], "input_dim": [[4, "cmdoption-arg-input_dim"]], "input_int32": [[4, "cmdoption-arg-input_int32"]], "input_layouts": [[4, "cmdoption-arg-input_layouts"]], "input_model_dtype": [[4, "cmdoption-arg-245"], [4, "cmdoption-arg-input_model_dtype"]], "input_name": [[4, "cmdoption-arg-input_name"]], "input_names": [[4, "cmdoption-arg-input_names"]], "input_nodes": [[4, "cmdoption-arg-input_nodes"]], "input_shape": [[4, "cmdoption-arg-input_shape"]], "input_shapes": [[4, "cmdoption-arg-input_shapes"]], "input_types": [[4, "cmdoption-arg-input_types"]], "inputs_to_make_channel_last": [[4, "cmdoption-arg-inputs_to_make_channel_last"]], "inside_layer_modules": [[4, "cmdoption-arg-inside_layer_modules"]], "int4_accuracy_level": [[4, "cmdoption-arg-int4_accuracy_level"]], "int4_block_size": [[4, "cmdoption-arg-int4_block_size"]], "inter_thread_num_list": [[4, "cmdoption-arg-inter_thread_num_list"]], "intra_thread_num_list": [[4, "cmdoption-arg-intra_thread_num_list"]], "io_bind": [[4, "cmdoption-arg-io_bind"]], "k_bits": [[4, "cmdoption-arg-k_bits"]], "k_clip_ratio": [[4, "cmdoption-arg-k_clip_ratio"]], "k_groupsize": [[4, "cmdoption-arg-k_groupsize"]], "keep_io_types": [[4, "cmdoption-arg-19"], [4, "cmdoption-arg-keep_io_types"]], "layer_name_filter": [[4, "cmdoption-arg-249"], [4, "cmdoption-arg-layer_name_filter"]], "layers_block_name": [[4, "cmdoption-arg-layers_block_name"]], "learning_rate (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.learning_rate"]], "lib_name": [[4, "cmdoption-arg-lib_name"]], "lib_targets": [[4, "cmdoption-arg-lib_targets"]], "loftq_iter": [[4, "cmdoption-arg-loftq_iter"]], "logger": [[4, "cmdoption-arg-logger"]], "lora_alpha": [[4, "cmdoption-arg-191"], [4, "cmdoption-arg-201"], [4, "cmdoption-arg-lora_alpha"]], "lora_dropout": [[4, "cmdoption-arg-192"], [4, "cmdoption-arg-202"], [4, "cmdoption-arg-lora_dropout"]], "lora_r": [[4, "cmdoption-arg-190"], [4, "cmdoption-arg-200"], [4, "cmdoption-arg-lora_r"]], "lr_scheduler_type (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.lr_scheduler_type"]], "make_inputs": [[4, "cmdoption-arg-make_inputs"]], "max_finite_val": [[4, "cmdoption-arg-max_finite_val"]], "max_layer": [[4, "cmdoption-arg-248"], [4, "cmdoption-arg-max_layer"]], "merge_adapter_weights": [[4, "cmdoption-arg-merge_adapter_weights"]], "metadata_only": [[4, "cmdoption-arg-metadata_only"]], "metric": [[4, "cmdoption-arg-122"], [4, "cmdoption-arg-142"], [4, "cmdoption-arg-metric"]], "min_layer": [[4, "cmdoption-arg-247"], [4, "cmdoption-arg-min_layer"]], "min_positive_val": [[4, "cmdoption-arg-min_positive_val"]], "model_type": [[4, "cmdoption-arg-217"], [4, "cmdoption-arg-model_type"]], "modules_to_fuse": [[4, "cmdoption-arg-modules_to_fuse"]], "modules_to_not_convert": [[4, "cmdoption-arg-modules_to_not_convert"]], "modules_to_save": [[4, "cmdoption-arg-193"], [4, "cmdoption-arg-203"], [4, "cmdoption-arg-modules_to_save"]], "name_pattern": [[4, "cmdoption-arg-name_pattern"]], "need_layer_fusing": [[4, "cmdoption-arg-need_layer_fusing"]], "no_repeat_ngram_size": [[4, "cmdoption-arg-no_repeat_ngram_size"]], "node_block_list": [[4, "cmdoption-arg-node_block_list"]], "nodes_to_exclude": [[4, "cmdoption-arg-158"], [4, "cmdoption-arg-56"], [4, "cmdoption-arg-78"], [4, "cmdoption-arg-nodes_to_exclude"]], "nodes_to_quantize": [[4, "cmdoption-arg-157"], [4, "cmdoption-arg-55"], [4, "cmdoption-arg-77"], [4, "cmdoption-arg-nodes_to_quantize"]], "nsamples": [[4, "cmdoption-arg-nsamples"]], "num_epochs": [[4, "cmdoption-arg-num_epochs"]], "num_heads": [[4, "cmdoption-arg-num_heads"]], "num_key_value_heads": [[4, "cmdoption-arg-num_key_value_heads"]], "num_steps": [[4, "cmdoption-arg-num_steps"]], "only_onnxruntime": [[4, "cmdoption-arg-only_onnxruntime"]], "op_block_list": [[4, "cmdoption-arg-30"], [4, "cmdoption-arg-op_block_list"]], "op_type_dict": [[4, "cmdoption-arg-124"], [4, "cmdoption-arg-144"], [4, "cmdoption-arg-op_type_dict"]], "op_types_to_quantize": [[4, "cmdoption-arg-156"], [4, "cmdoption-arg-53"], [4, "cmdoption-arg-75"], [4, "cmdoption-arg-op_types_to_quantize"]], "opt_level": [[4, "cmdoption-arg-opt_level"]], "opt_level_list": [[4, "cmdoption-arg-opt_level_list"]], "optim (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.optim"]], "optimization_options": [[4, "cmdoption-arg-optimization_options"]], "optimize_model": [[4, "cmdoption-arg-optimize_model"]], "optional_inputs": [[4, "cmdoption-arg-optional_inputs"]], "out_node": [[4, "cmdoption-arg-out_node"]], "output_dir (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.output_dir"]], "output_model": [[4, "cmdoption-arg-output_model"]], "output_names": [[4, "cmdoption-arg-output_names"]], "output_nodes": [[4, "cmdoption-arg-output_nodes"]], "outputs_to_make_channel_last": [[4, "cmdoption-arg-outputs_to_make_channel_last"]], "outside_layer_modules": [[4, "cmdoption-arg-outside_layer_modules"]], "overrides_config": [[4, "cmdoption-arg-overrides_config"]], "overwrite_output_dir (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.overwrite_output_dir"]], "pack_inputs": [[4, "cmdoption-arg-pack_inputs"]], "pack_model_for_onnx_conversion": [[4, "cmdoption-arg-pack_model_for_onnx_conversion"]], "parallel_jobs": [[4, "cmdoption-arg-parallel_jobs"]], "past_key_value_name": [[4, "cmdoption-arg-past_key_value_name"]], "per_channel": [[4, "cmdoption-arg-159"], [4, "cmdoption-arg-57"], [4, "cmdoption-arg-79"], [4, "cmdoption-arg-per_channel"]], "percdamp": [[4, "cmdoption-arg-percdamp"]], "post": [[4, "cmdoption-arg-post"]], "pre": [[4, "cmdoption-arg-pre"]], "precision": [[4, "cmdoption-arg-precision"]], "prepare_qnn_config": [[4, "cmdoption-arg-86"], [4, "cmdoption-arg-prepare_qnn_config"]], "preset": [[4, "cmdoption-arg-preset"]], "providers_list": [[4, "cmdoption-arg-providers_list"]], "ptl_data_module": [[4, "cmdoption-arg-ptl_data_module"]], "ptl_module": [[4, "cmdoption-arg-ptl_module"]], "q_group_size": [[4, "cmdoption-arg-q_group_size"]], "qconfig_func": [[4, "cmdoption-arg-qconfig_func"]], "qnn_extra_options": [[4, "cmdoption-arg-87"], [4, "cmdoption-arg-qnn_extra_options"]], "quant_format": [[4, "cmdoption-arg-126"], [4, "cmdoption-arg-146"], [4, "cmdoption-arg-162"], [4, "cmdoption-arg-84"], [4, "cmdoption-arg-quant_format"]], "quant_level": [[4, "cmdoption-arg-119"], [4, "cmdoption-arg-139"], [4, "cmdoption-arg-quant_level"]], "quant_mode": [[4, "cmdoption-arg-153"], [4, "cmdoption-arg-51"], [4, "cmdoption-arg-73"], [4, "cmdoption-arg-quant_mode"]], "quant_preprocess": [[4, "cmdoption-arg-160"], [4, "cmdoption-arg-59"], [4, "cmdoption-arg-81"], [4, "cmdoption-arg-quant_preprocess"]], "quant_type": [[4, "cmdoption-arg-quant_type"]], "recipes": [[4, "cmdoption-arg-117"], [4, "cmdoption-arg-137"], [4, "cmdoption-arg-recipes"]], "reduce_range": [[4, "cmdoption-arg-106"], [4, "cmdoption-arg-118"], [4, "cmdoption-arg-138"], [4, "cmdoption-arg-58"], [4, "cmdoption-arg-80"], [4, "cmdoption-arg-reduce_range"]], "report_to (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.report_to"]], "resume_from_checkpoint (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.resume_from_checkpoint"]], "rotate": [[4, "cmdoption-arg-rotate"]], "rotation_seed": [[4, "cmdoption-arg-rotation_seed"]], "round_interval": [[4, "cmdoption-arg-round_interval"]], "save_as_external_data": [[4, "cmdoption-arg-1"], [4, "cmdoption-arg-100"], [4, "cmdoption-arg-107"], [4, "cmdoption-arg-11"], [4, "cmdoption-arg-127"], [4, "cmdoption-arg-148"], [4, "cmdoption-arg-167"], [4, "cmdoption-arg-173"], [4, "cmdoption-arg-179"], [4, "cmdoption-arg-184"], [4, "cmdoption-arg-20"], [4, "cmdoption-arg-223"], [4, "cmdoption-arg-25"], [4, "cmdoption-arg-258"], [4, "cmdoption-arg-31"], [4, "cmdoption-arg-36"], [4, "cmdoption-arg-41"], [4, "cmdoption-arg-46"], [4, "cmdoption-arg-6"], [4, "cmdoption-arg-68"], [4, "cmdoption-arg-95"], [4, "cmdoption-arg-save_as_external_data"]], "save_format": [[4, "cmdoption-arg-save_format"]], "save_metadata_for_token_generation": [[4, "cmdoption-arg-save_metadata_for_token_generation"]], "script_dir": [[4, "cmdoption-arg-17"], [4, "cmdoption-arg-210"], [4, "cmdoption-arg-213"], [4, "cmdoption-arg-215"], [4, "cmdoption-arg-240"], [4, "cmdoption-arg-244"], [4, "cmdoption-arg-253"], [4, "cmdoption-arg-script_dir"]], "search": [[4, "cmdoption-arg-search"]], "seed": [[4, "cmdoption-arg-233"], [4, "cmdoption-arg-238"], [4, "cmdoption-arg-241"], [4, "cmdoption-arg-seed"], [6, "cmdoption-arg-1"], [6, "cmdoption-arg-seed"]], "size_threshold": [[4, "cmdoption-arg-103"], [4, "cmdoption-arg-110"], [4, "cmdoption-arg-130"], [4, "cmdoption-arg-14"], [4, "cmdoption-arg-151"], [4, "cmdoption-arg-170"], [4, "cmdoption-arg-176"], [4, "cmdoption-arg-182"], [4, "cmdoption-arg-187"], [4, "cmdoption-arg-226"], [4, "cmdoption-arg-23"], [4, "cmdoption-arg-261"], [4, "cmdoption-arg-28"], [4, "cmdoption-arg-34"], [4, "cmdoption-arg-39"], [4, "cmdoption-arg-4"], [4, "cmdoption-arg-44"], [4, "cmdoption-arg-49"], [4, "cmdoption-arg-71"], [4, "cmdoption-arg-9"], [4, "cmdoption-arg-98"], [4, "cmdoption-arg-size_threshold"]], "sparsity": [[4, "cmdoption-arg-234"], [4, "cmdoption-arg-sparsity"]], "static_groups": [[4, "cmdoption-arg-static_groups"]], "strict": [[4, "cmdoption-arg-strict"]], "sym": [[4, "cmdoption-arg-sym"]], "target_device": [[4, "cmdoption-arg-221"], [4, "cmdoption-arg-target_device"]], "target_modules": [[4, "cmdoption-arg-target_modules"]], "target_opset": [[4, "cmdoption-arg-0"], [4, "cmdoption-arg-172"], [4, "cmdoption-arg-222"], [4, "cmdoption-arg-254"], [4, "cmdoption-arg-target_opset"]], "tool_command": [[4, "cmdoption-arg-tool_command"]], "tool_command_args": [[4, "cmdoption-arg-tool_command_args"]], "torch_dtype": [[4, "cmdoption-arg-189"], [4, "cmdoption-arg-194"], [4, "cmdoption-arg-204"], [4, "cmdoption-arg-torch_dtype"]], "train_data_config": [[4, "cmdoption-arg-196"], [4, "cmdoption-arg-206"], [4, "cmdoption-arg-211"], [4, "cmdoption-arg-train_data_config"]], "training_args": [[4, "cmdoption-arg-198"], [4, "cmdoption-arg-208"], [4, "cmdoption-arg-training_args"]], "training_loop_func": [[4, "cmdoption-arg-training_loop_func"]], "trt_fp16_enable": [[4, "cmdoption-arg-trt_fp16_enable"]], "true_sequential": [[4, "cmdoption-arg-true_sequential"]], "tuning_criterion": [[4, "cmdoption-arg-121"], [4, "cmdoption-arg-141"], [4, "cmdoption-arg-tuning_criterion"]], "use_dynamo_exporter": [[4, "cmdoption-arg-use_dynamo_exporter"]], "use_enhanced_quantizer": [[4, "cmdoption-arg-use_enhanced_quantizer"]], "use_external_data_format": [[4, "cmdoption-arg-use_external_data_format"]], "use_forced_decoder_ids": [[4, "cmdoption-arg-use_forced_decoder_ids"]], "use_gpu": [[4, "cmdoption-arg-178"], [4, "cmdoption-arg-use_gpu"]], "use_gqa": [[4, "cmdoption-arg-use_gqa"]], "use_logits_processor": [[4, "cmdoption-arg-use_logits_processor"]], "use_prefix_vocab_mask": [[4, "cmdoption-arg-use_prefix_vocab_mask"]], "use_symbolic_shape_infer": [[4, "cmdoption-arg-use_symbolic_shape_infer"]], "use_temperature": [[4, "cmdoption-arg-use_temperature"]], "use_vocab_mask": [[4, "cmdoption-arg-use_vocab_mask"]], "user_script": [[4, "cmdoption-arg-16"], [4, "cmdoption-arg-209"], [4, "cmdoption-arg-212"], [4, "cmdoption-arg-214"], [4, "cmdoption-arg-239"], [4, "cmdoption-arg-243"], [4, "cmdoption-arg-252"], [4, "cmdoption-arg-user_script"]], "v_bits": [[4, "cmdoption-arg-v_bits"]], "v_clip_ratio": [[4, "cmdoption-arg-v_clip_ratio"]], "v_groupsize": [[4, "cmdoption-arg-v_groupsize"]], "val_data_config": [[4, "cmdoption-arg-val_data_config"]], "version": [[4, "cmdoption-arg-version"], [5, "cmdoption-arg-version"]], "w_asym": [[4, "cmdoption-arg-w_asym"]], "w_bit": [[4, "cmdoption-arg-w_bit"]], "w_bits": [[4, "cmdoption-arg-w_bits"]], "w_gptq": [[4, "cmdoption-arg-w_gptq"]], "w_groupsize": [[4, "cmdoption-arg-w_groupsize"]], "w_rtn": [[4, "cmdoption-arg-w_rtn"]], "warmup_ratio (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.warmup_ratio"]], "weight_only_config": [[4, "cmdoption-arg-123"], [4, "cmdoption-arg-143"], [4, "cmdoption-arg-weight_only_config"]], "weight_type": [[4, "cmdoption-arg-155"], [4, "cmdoption-arg-52"], [4, "cmdoption-arg-74"], [4, "cmdoption-arg-weight_type"]], "workspace": [[4, "cmdoption-arg-116"], [4, "cmdoption-arg-136"], [4, "cmdoption-arg-workspace"]], "zero_point": [[4, "cmdoption-arg-zero_point"]], "azuremldatastore (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.AzureMLDatastore"]], "azuremljoboutput (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.AzureMLJobOutput"]], "azuremlmodel (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.AzureMLModel"]], "azuremlregistrymodel (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.AzureMLRegistryModel"]], "localfile (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.LocalFile"]], "localfolder (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.LocalFolder"]], "resourcetype (class in olive.resource_path)": [[5, "olive.resource_path.ResourceType"]], "stringname (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.StringName"]], "aml_config_path (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.aml_config_path"]], "azureml_client": [[5, "cmdoption-arg-2"], [5, "cmdoption-arg-3"], [5, "cmdoption-arg-azureml_client"]], "create_client() (olive.azureml.azureml_client.azuremlclientconfig method)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.create_client"]], "create_registry_client() (olive.azureml.azureml_client.azuremlclientconfig method)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.create_registry_client"]], "datastore_name": [[5, "cmdoption-arg-datastore_name"]], "datastore_url": [[5, "cmdoption-arg-datastore_url"]], "default_auth_params (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.default_auth_params"]], "get_workspace_config() (olive.azureml.azureml_client.azuremlclientconfig method)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.get_workspace_config"]], "job_name": [[5, "cmdoption-arg-job_name"]], "keyvault_name (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.keyvault_name"]], "max_operation_retries (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.max_operation_retries"]], "name": [[5, "cmdoption-arg-1"], [5, "cmdoption-arg-name"]], "operation_retry_interval (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.operation_retry_interval"]], "output_name": [[5, "cmdoption-arg-output_name"]], "path": [[5, "cmdoption-arg-0"], [5, "cmdoption-arg-path"]], "read_timeout (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.read_timeout"]], "relative_path": [[5, "cmdoption-arg-4"], [5, "cmdoption-arg-relative_path"]], "resource_group (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.resource_group"]], "subscription_id (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.subscription_id"]], "workspace_name (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.workspace_name"]], "group": [[6, "cmdoption-arg-group"]], "multivariate": [[6, "cmdoption-arg-multivariate"]], "num_samples": [[6, "cmdoption-arg-0"], [6, "cmdoption-arg-num_samples"]], "with_replacement": [[6, "cmdoption-arg-with_replacement"]], "acceleratorconfig (class in olive.systems.common)": [[7, "olive.systems.common.AcceleratorConfig"]], "azureml (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.AzureML"]], "azuremlsystem (class in olive.systems.azureml)": [[7, "olive.systems.azureml.AzureMLSystem"]], "azurend12ssystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND12SSystem"]], "azurend24rssystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND24RSSystem"]], "azurend24ssystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND24SSystem"]], "azurend6ssystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND6SSystem"]], "azurend96a100system (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND96A100System"]], "azurend96asystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND96ASystem"]], "azurendv2system (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureNDV2System"]], "cpu (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.CPU"]], "cpu_spr (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.CPU_SPR"]], "device (class in olive.hardware.accelerator)": [[7, "olive.hardware.accelerator.Device"]], "docker (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.Docker"]], "dockersystem (class in olive.systems.docker)": [[7, "olive.systems.docker.DockerSystem"]], "gpu (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.GPU"]], "intel_myriad (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.INTEL_MYRIAD"]], "isolatedort (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.IsolatedORT"]], "isolatedortsystem (class in olive.systems.isolated_ort)": [[7, "olive.systems.isolated_ort.IsolatedORTSystem"]], "local (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.Local"]], "localsystem (class in olive.systems.local)": [[7, "olive.systems.local.LocalSystem"]], "npu (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.NPU"]], "pythonenvironment (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.PythonEnvironment"]], "pythonenvironmentsystem (class in olive.systems.python_environment)": [[7, "olive.systems.python_environment.PythonEnvironmentSystem"]], "surfaceprosystem1796 (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.SurfaceProSystem1796"]], "surfaceprosystem1807 (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.SurfaceProSystem1807"]], "systemtype (class in olive.systems.common)": [[7, "olive.systems.common.SystemType"]], "vpu (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.VPU"]], "accelerators (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.accelerators"]], "accelerators (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.accelerators"]], "accelerators (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.accelerators"]], "accelerators (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.accelerators"]], "accelerators (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.accelerators"]], "accelerators (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.accelerators"]], "accelerators (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.accelerators"]], "accelerators (olive.systems.system_alias.surfaceprosystem1796 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1796.accelerators"]], "accelerators (olive.systems.system_alias.surfaceprosystem1807 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1807.accelerators"]], "accelerators (olive.systems.system_config.localtargetuserconfig attribute)": [[7, "olive.systems.system_config.LocalTargetUserConfig.accelerators"]], "aml_compute (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.aml_compute"]], "aml_docker_config (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.aml_docker_config"]], "aml_environment_config (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.aml_environment_config"]], "azureml_client_config (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.azureml_client_config"]], "base_image (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.base_image"]], "build_args (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.build_args"]], "build_context_path (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.build_context_path"]], "build_context_path (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.build_context_path"]], "conda_file_path (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.conda_file_path"]], "config (olive.systems.system_config.systemconfig attribute)": [[7, "olive.systems.system_config.SystemConfig.config"]], "create_system() (olive.systems.system_config.systemconfig method)": [[7, "olive.systems.system_config.SystemConfig.create_system"]], "datastores (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.datastores"]], "device (olive.systems.common.acceleratorconfig attribute)": [[7, "olive.systems.common.AcceleratorConfig.device"]], "dockerfile (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.dockerfile"]], "dockerfile (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.dockerfile"]], "environment_variables (olive.systems.system_config.isolatedorttargetuserconfig attribute)": [[7, "olive.systems.system_config.IsolatedORTTargetUserConfig.environment_variables"]], "execution_providers (olive.systems.common.acceleratorconfig attribute)": [[7, "olive.systems.common.AcceleratorConfig.execution_providers"]], "hf_token (olive.systems.system_config.localtargetuserconfig attribute)": [[7, "olive.systems.system_config.LocalTargetUserConfig.hf_token"]], "image_name (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.image_name"]], "instance_count (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.instance_count"]], "is_dev (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.is_dev"]], "is_dev (olive.systems.system_config.dockertargetuserconfig attribute)": [[7, "olive.systems.system_config.DockerTargetUserConfig.is_dev"]], "local_docker_config (olive.systems.system_config.dockertargetuserconfig attribute)": [[7, "olive.systems.system_config.DockerTargetUserConfig.local_docker_config"]], "module": [[7, "module-olive.systems.system_alias"]], "name (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.name"]], "num_cpus (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.num_cpus"]], "num_gpus (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.num_gpus"]], "olive.systems.system_alias": [[7, "module-olive.systems.system_alias"]], "olive_managed_env (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.olive_managed_env"]], "olive_managed_env (olive.systems.system_config.dockertargetuserconfig attribute)": [[7, "olive.systems.system_config.DockerTargetUserConfig.olive_managed_env"]], "olive_managed_env (olive.systems.system_config.pythonenvironmenttargetuserconfig attribute)": [[7, "olive.systems.system_config.PythonEnvironmentTargetUserConfig.olive_managed_env"]], "olive_managed_env (olive.systems.system_config.systemconfig property)": [[7, "olive.systems.system_config.SystemConfig.olive_managed_env"]], "prepend_to_path (olive.systems.system_config.isolatedorttargetuserconfig attribute)": [[7, "olive.systems.system_config.IsolatedORTTargetUserConfig.prepend_to_path"]], "python_environment_path (olive.systems.system_config.isolatedorttargetuserconfig attribute)": [[7, "olive.systems.system_config.IsolatedORTTargetUserConfig.python_environment_path"]], "requirements_file (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.requirements_file"]], "requirements_file (olive.systems.system_config.dockertargetuserconfig attribute)": [[7, "olive.systems.system_config.DockerTargetUserConfig.requirements_file"]], "requirements_file (olive.systems.system_config.pythonenvironmenttargetuserconfig attribute)": [[7, "olive.systems.system_config.PythonEnvironmentTargetUserConfig.requirements_file"]], "resources (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.resources"]], "run_params (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.run_params"]], "sku (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.sku"]], "sku (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.sku"]], "sku (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.sku"]], "sku (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.sku"]], "sku (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.sku"]], "sku (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.sku"]], "sku (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.sku"]], "sku (olive.systems.system_alias.surfaceprosystem1796 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1796.sku"]], "sku (olive.systems.system_alias.surfaceprosystem1807 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1807.sku"]], "system_type (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.system_type"]], "system_type (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.system_type"]], "system_type (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.system_type"]], "system_type (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.system_type"]], "system_type (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.system_type"]], "system_type (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.system_type"]], "system_type (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.system_type"]], "system_type (olive.systems.system_alias.surfaceprosystem1796 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1796.system_type"]], "system_type (olive.systems.system_alias.surfaceprosystem1807 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1807.system_type"]], "tags (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.tags"]], "type (olive.systems.system_config.systemconfig attribute)": [[7, "olive.systems.system_config.SystemConfig.type"]], "version (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.version"]]}})
\ No newline at end of file
+Search.setIndex({"docnames": ["api/engine", "api/evaluator", "api/metric", "api/models", "api/passes", "api/resource_path", "api/search-algorithms", "api/systems", "examples", "extending_olive/custom_scripts", "extending_olive/design", "extending_olive/how_to_add_optimization_pass", "features/azureml_integration", "features/cli", "features/cloud_model_cache", "features/conversion", "features/huggingface_model_optimization", "features/lora", "features/model_transformations_and_optimizations", "features/packaging_output_models", "features/passes/convert_onnx", "features/passes/convert_pytorch", "features/passes/onnx", "features/passes/openvino", "features/passes/pytorch", "features/passes/qnn", "features/passes/quant_onnx", "features/passes/quant_pytorch", "features/passes/snpe", "features/quantization", "features/run_workflow_remotely", "getstarted/installation", "getstarted/quickstart_bert_example", "getstarted/quickstart_examples", "getstarted/quickstart_resnet_example", "index", "overview/olive", "overview/options", "overview/quicktour", "tutorials/advanced_users", "tutorials/azure_arc", "tutorials/azureml_scripts", "tutorials/configure_auto_optimizer", "tutorials/configure_data", "tutorials/configure_metrics", "tutorials/configure_model_path", "tutorials/configure_pass", "tutorials/configure_systems"], "filenames": ["api/engine.rst", "api/evaluator.rst", "api/metric.rst", "api/models.rst", "api/passes.rst", "api/resource_path.rst", "api/search-algorithms.rst", "api/systems.rst", "examples.md", "extending_olive/custom_scripts.md", "extending_olive/design.md", "extending_olive/how_to_add_optimization_pass.md", "features/azureml_integration.md", "features/cli.rst", "features/cloud_model_cache.md", "features/conversion.rst", "features/huggingface_model_optimization.md", "features/lora.md", "features/model_transformations_and_optimizations.rst", "features/packaging_output_models.md", "features/passes/convert_onnx.md", "features/passes/convert_pytorch.md", "features/passes/onnx.md", "features/passes/openvino.md", "features/passes/pytorch.md", "features/passes/qnn.md", "features/passes/quant_onnx.md", "features/passes/quant_pytorch.md", "features/passes/snpe.md", "features/quantization.rst", "features/run_workflow_remotely.md", "getstarted/installation.md", "getstarted/quickstart_bert_example.md", "getstarted/quickstart_examples.rst", "getstarted/quickstart_resnet_example.md", "index.rst", "overview/olive.md", "overview/options.md", "overview/quicktour.md", "tutorials/advanced_users.md", "tutorials/azure_arc.md", "tutorials/azureml_scripts.md", "tutorials/configure_auto_optimizer.rst", "tutorials/configure_data.rst", "tutorials/configure_metrics.rst", "tutorials/configure_model_path.md", "tutorials/configure_pass.rst", "tutorials/configure_systems.rst"], "titles": ["Engine", "OliveEvaluator", "Metric", "OliveModels", "Passes", "ResourcePath", "SearchAlgorithms", "OliveSystems", "Examples", "Custom Scripts", "Design", "How to add new optimization Pass", "Azure ML integration", "Command Line Tools", "Cloud Model Cache", "Model Conversions", "Huggingface Integration", "LoRA Adapters", "Model Transformations and Optimizations", "Packaging Olive artifacts", "ONNX", "PyTorch", "ONNX", "OpenVINO", "PyTorch", "QNN", "ONNX", "PyTorch", "SNPE", "Model Quantizations", "Remote Workflow", "Installation", "BERT optimization with PTQ on CPU", "Quickstart Examples", "ResNet optimization with PTQ on CPU", "Welcome to Olive\u2019s documentation!", "Olive", "Olive Options", "Quick Tour", "Advanced User Tour", "Self-hosted Kubernetes cluster", "Azure ML scripts", "How To Configure Auto Optimizer", "How To Configure Data", "How To Configure Metric", "How To Set Model Path", "How To Configure Pass", "How To Configure System"], "terms": {"class": [0, 1, 2, 3, 4, 5, 9, 37, 43, 44, 46, 47], "oliv": [0, 1, 2, 3, 4, 5, 6, 7, 9, 10, 11, 14, 15, 16, 17, 18, 22, 24, 26, 27, 29, 31, 32, 34, 39, 41, 42, 43, 44, 45, 46, 47], "workflow_id": [0, 30, 37], "str": [0, 2, 3, 4, 5, 7, 9, 11, 14, 19, 37, 42, 43], "default_workflow": 0, "search_strategi": [0, 19, 37, 39, 42], "dict": [0, 2, 3, 4, 5, 7, 9, 11, 19, 37], "ani": [0, 2, 3, 4, 5, 7, 10, 11, 26, 37, 38, 43, 46], "searchstrategyconfig": 0, "none": [0, 2, 3, 4, 5, 7, 19, 37, 42, 43, 44, 47], "host": [0, 10, 11, 16, 19, 35, 42, 47], "systemconfig": [0, 7], "target": [0, 4, 8, 10, 13, 16, 19, 20, 25, 26, 36, 37, 38, 42, 44, 47], "evalu": [0, 1, 2, 3, 4, 9, 12, 13, 14, 16, 19, 42, 43, 44, 47], "oliveevaluatorconfig": [0, 39], "cache_dir": [0, 19, 37, 39, 42], "cach": [0, 4, 10, 19, 25, 30, 35, 37, 39, 42], "clean_cach": [0, 37], "bool": [0, 3, 4, 6, 7, 14, 19, 37, 42], "fals": [0, 3, 4, 6, 7, 11, 13, 14, 16, 19, 22, 37, 39, 42, 46, 47], "clean_evaluation_cach": [0, 37], "plot_pareto_fronti": [0, 37], "azureml_client_config": [0, 7, 47], "sourc": [0, 1, 2, 3, 4, 5, 7], "The": [0, 3, 4, 5, 6, 7, 10, 11, 13, 14, 16, 17, 18, 19, 20, 22, 23, 24, 25, 26, 28, 30, 36, 37, 38, 39, 41, 42, 43, 45, 47], "execut": [0, 3, 4, 11, 13, 14, 19, 22, 25, 26, 28, 37, 42, 47], "regist": [0, 10, 19, 37, 45], "step": [0, 4, 11, 14, 16, 22, 26, 30, 37, 38, 40], "It": [0, 4, 10, 11, 13, 16, 19, 22, 23, 24, 27, 30, 37, 38, 40, 46, 47], "facilit": 0, "output": [0, 4, 5, 9, 10, 11, 13, 14, 15, 17, 18, 19, 22, 24, 32, 34, 36, 37, 38, 39, 42, 43], "model": [0, 4, 5, 8, 10, 11, 17, 21, 24, 26, 27, 35, 36, 42, 43, 44, 47], "us": [0, 3, 4, 5, 6, 9, 10, 11, 13, 16, 17, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 31, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47], "provid": [0, 3, 4, 5, 9, 10, 11, 16, 17, 18, 19, 22, 24, 25, 26, 28, 30, 35, 36, 37, 38, 39, 41, 42, 43, 44, 46, 47], "criteria": 0, "produc": [0, 17, 18, 38], "": [0, 4, 9, 16, 17, 21, 36, 37, 38, 39, 43, 46], "pass_typ": 0, "type": [0, 2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 13, 16, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 37, 38, 39, 42, 43, 45, 46, 47], "pass": [0, 9, 12, 14, 16, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 32, 34, 35, 40, 42, 43, 47], "config": [0, 3, 4, 5, 10, 11, 12, 17, 20, 22, 23, 24, 25, 26, 27, 28, 30, 31, 38, 39, 40, 41, 42, 44, 45, 46, 47], "disable_search": [0, 11, 26, 37, 46], "name": [0, 2, 4, 5, 6, 7, 9, 11, 12, 14, 16, 19, 22, 23, 30, 37, 38, 39, 41, 42, 43, 44, 47], "olivesystem": [0, 35, 39], "evaluator_config": [0, 39], "clean_run_cach": [0, 37], "output_nam": [0, 3, 4, 5, 12, 16, 28, 37, 38, 39, 45], "configur": [0, 4, 5, 6, 9, 10, 12, 16, 17, 35, 37, 38, 39, 45], "so": [0, 4, 11, 14, 19, 20, 22, 25], "could": [0, 4, 11, 18, 37, 38, 47], "instanti": [0, 39], "later": [0, 25], "run": [0, 4, 10, 12, 14, 16, 17, 19, 20, 22, 23, 24, 25, 26, 28, 37, 38, 40, 42, 43, 44, 47], "input_model_config": 0, "modelconfig": [0, 3, 39], "accelerator_spec": [0, 11, 19], "list": [0, 2, 3, 4, 5, 6, 7, 10, 13, 16, 19, 22, 31, 37, 38, 39, 43, 46, 47], "acceleratorspec": [0, 11], "packaging_config": [0, 19, 37], "packagingconfig": [0, 19, 37], "output_dir": [0, 4, 19, 37, 42], "evaluate_input_model": [0, 37], "true": [0, 2, 4, 5, 6, 11, 13, 14, 16, 19, 20, 22, 26, 27, 28, 37, 42, 44, 46, 47], "log_to_fil": [0, 37], "log_severity_level": [0, 37, 38], "int": [0, 3, 4, 5, 6, 7, 11, 19, 37, 42], "1": [0, 4, 6, 7, 9, 12, 13, 16, 19, 20, 22, 26, 28, 37, 38, 39, 42, 43, 44, 45, 47], "cloud_cache_config": [0, 14], "cloudcacheconfig": 0, "all": [0, 3, 4, 8, 10, 13, 19, 20, 36, 37, 38, 39, 42, 43, 46, 47], "input": [0, 3, 4, 10, 11, 12, 13, 14, 15, 17, 18, 19, 22, 23, 24, 26, 28, 42, 43, 47], "one": [0, 4, 8, 10, 11, 15, 19, 22, 24, 28, 38, 39, 41, 44], "more": [0, 4, 5, 6, 10, 12, 14, 16, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 37, 38, 39, 41, 42, 43, 47], "candid": [0, 10, 19, 32, 34], "paramet": [0, 4, 5, 6, 10, 11, 13, 17, 20, 22, 23, 24, 25, 26, 27, 28, 37, 39, 43, 46], "acceler": [0, 4, 7, 11, 19, 20, 21, 22, 24, 25, 26, 28, 36, 37, 42], "spec": [0, 10, 47], "packag": [0, 10, 11, 13, 27, 31, 32, 34, 35, 37, 38, 47], "i": [0, 3, 4, 5, 6, 9, 10, 11, 13, 16, 17, 18, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 31, 32, 34, 35, 36, 37, 38, 40, 41, 42, 43, 44, 45, 46, 47], "zip": [0, 19, 28], "file": [0, 4, 5, 9, 11, 12, 13, 16, 17, 22, 24, 25, 26, 27, 28, 30, 37, 38, 39, 40, 41, 42, 47], "directori": [0, 4, 9, 10, 13, 16, 28, 32, 34, 37, 38, 43, 44, 45], "save": [0, 4, 13, 22, 32, 34, 37, 45], "prefix": [0, 19, 37], "log": [0, 4, 14, 16, 30, 37], "sever": [0, 12, 37, 40, 43], "level": [0, 4, 13, 20, 30, 37, 38, 45], "logger": [0, 4], "cloud": [0, 35, 36], "return": [0, 4, 9, 10, 11, 37, 39, 44], "search": [0, 4, 6, 11, 13, 18, 37, 39, 42, 46], "strategi": [0, 4, 16, 37], "ar": [0, 3, 4, 5, 6, 7, 10, 11, 12, 14, 16, 18, 19, 20, 22, 23, 24, 26, 36, 37, 38, 39, 41, 47], "order": [0, 4, 19, 22, 36, 37, 39], "thei": [0, 13, 19, 37, 43, 47], "were": [0, 4, 19, 37], "final": [0, 4, 10, 11, 14, 37, 38, 39, 47], "_": [0, 19, 43], "_model": 0, "onnx": [0, 4, 8, 10, 13, 15, 17, 18, 19, 25, 27, 28, 29, 32, 34, 37, 38, 39, 42, 47], "json": [0, 4, 9, 12, 13, 17, 30, 34, 38, 39, 41, 42, 43, 44, 46, 47], "result": [0, 4, 20, 24, 37, 39, 44, 47], "_metric": 0, "footprint": [0, 23, 24, 26], "find": [0, 10, 12, 16, 20, 22, 24, 30, 37, 38], "configbas": [0, 2], "can": [0, 4, 9, 10, 11, 12, 13, 14, 16, 17, 18, 19, 20, 21, 22, 24, 25, 26, 27, 28, 30, 31, 32, 34, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47], "assign": [0, 4, 46], "dictionari": [0, 4, 10, 37, 38, 43, 46, 47], "kei": [0, 4, 9, 13, 16, 37, 43], "correspond": [0, 4, 13, 14, 22, 28, 32, 34], "field": [0, 2, 4, 5, 7, 11, 13, 42, 43, 44], "pydant": [0, 2, 4, 5, 7], "set": [0, 2, 4, 5, 7, 10, 11, 13, 14, 16, 19, 20, 22, 25, 26, 28, 35, 36, 37, 38, 39, 40, 42, 46], "path": [0, 3, 4, 5, 7, 9, 11, 12, 13, 16, 19, 25, 28, 30, 35, 37, 38, 40, 41, 43, 47], "execution_ord": [0, 10, 19, 37, 39, 42], "requir": [0, 2, 4, 5, 7, 10, 11, 13, 16, 19, 22, 24, 26, 27, 36, 37, 38, 39, 41, 42, 43, 45, 46, 47], "search_algorithm": [0, 10, 19, 37, 39, 42], "search_algorithm_config": [0, 19, 37, 42], "output_model_num": [0, 37], "stop_when_goals_met": [0, 37], "max_it": [0, 37], "max_tim": [0, 37], "system": [0, 11, 12, 14, 35, 38, 42], "system_config": [0, 7], "systemtyp": 0, "targetuserconfig": [0, 7], "create_system": [0, 7], "properti": [0, 7], "olive_managed_env": [0, 7, 47], "olive_evalu": [0, 1, 39], "type_arg": 0, "option": [0, 4, 5, 6, 10, 11, 12, 20, 22, 25, 35, 38, 39, 42, 43, 46, 47], "user_script": [0, 4, 22, 24, 26, 28, 37, 43, 44], "script_dir": [0, 3, 4, 37, 43], "metric": [0, 4, 9, 10, 35, 37, 38, 39, 42], "is_accuracy_drop_toler": 0, "create_evalu": 0, "olivemodelhandl": [0, 3, 4, 11, 39], "oliveevalu": [0, 35], "backend": [2, 4, 16, 25, 37, 42], "torch_metr": [2, 37], "sub_typ": [2, 9, 16, 37, 39, 42, 44], "submetr": 2, "user_config": [2, 37, 44], "data_config": [2, 4, 9, 16, 22, 23, 26, 27, 28, 37, 39, 42, 43, 44, 46], "dataconfig": [2, 4, 43, 44], "get_inference_set": 2, "framework": [2, 4, 23, 24, 25, 26, 28, 36, 39], "get_run_kwarg": 2, "get_sub_type_info": 2, "info_nam": 2, "no_priority_filt": 2, "callback": 2, "function": [2, 3, 4, 9, 35, 37, 39, 43, 44], "lambda": 2, "valu": [2, 4, 5, 7, 9, 10, 11, 13, 16, 19, 20, 22, 24, 26, 37, 39, 41, 42, 43, 44, 46], "an": [2, 3, 4, 5, 7, 8, 10, 11, 13, 14, 16, 17, 18, 19, 20, 21, 22, 24, 26, 27, 28, 36, 37, 38, 39, 41, 43, 44, 45, 47], "enumer": [2, 5, 7], "accuraci": [2, 4, 9, 10, 16, 19, 20, 26, 27, 36, 37, 38, 42], "custom": [2, 4, 8, 10, 13, 22, 24, 35, 37], "latenc": [2, 3, 4, 10, 19, 23, 36, 37, 38, 39, 42], "throughput": [2, 10, 36, 37], "accuracy_scor": [2, 9, 37, 44], "auroc": [2, 9, 37, 44], "f1_score": [2, 9, 37, 44], "perplex": [2, 4, 24], "precis": [2, 4, 13, 15, 21, 23, 26, 37, 42], "recal": [2, 37], "avg": [2, 19, 37, 39, 42, 44], "max": [2, 4, 5, 9, 13, 37, 42, 44], "min": [2, 4, 37, 42, 44], "p50": 2, "p75": 2, "p90": 2, "p95": 2, "p99": 2, "p999": 2, "float": [2, 4, 13, 23, 26, 37], "has_regression_go": 2, "follow": [3, 4, 5, 6, 7, 9, 10, 11, 12, 16, 19, 22, 28, 30, 31, 36, 37, 38, 40, 42, 43, 44, 47], "avail": [3, 4, 5, 6, 7, 11, 12, 13, 14, 16, 20, 25, 28, 31, 37, 39, 42, 43, 46, 47], "which": [3, 4, 10, 11, 16, 17, 19, 20, 22, 23, 24, 25, 26, 27, 37, 38, 39, 42, 43, 47], "creat": [3, 4, 5, 9, 10, 11, 13, 16, 19, 22, 27, 31, 37, 39, 41, 43, 47], "hfmodelhandl": [3, 4, 37], "model_path": [3, 12, 16, 19, 37, 38, 39, 42, 43, 45], "resourcepathconfig": 3, "resourcepath": [3, 35], "task": [3, 9, 13, 16, 22, 37, 38, 42, 43], "text": [3, 13, 16, 37, 42, 43], "gener": [3, 4, 6, 13, 16, 17, 18, 19, 20, 22, 28, 37, 38, 45, 47], "past": [3, 4, 13, 16, 37], "load_kwarg": [3, 37], "hfloadkwarg": 3, "io_config": [3, 16, 37, 38, 39], "ioconfig": 3, "adapter_path": [3, 13, 17], "model_attribut": [3, 19], "distributedhfmodelhandl": [3, 4], "model_name_pattern": 3, "num_rank": 3, "pytorchmodelhandl": [3, 4, 23, 37], "model_file_format": 3, "modelfileformat": 3, "pytorch_entire_model": 3, "model_load": 3, "callabl": [3, 4, 11], "model_script": 3, "dummy_inputs_func": 3, "besid": [3, 20, 43], "load": [3, 4, 9, 13, 22, 24, 25, 28, 37, 38, 39, 43, 44], "also": [3, 4, 10, 11, 13, 16, 18, 19, 20, 22, 24, 25, 26, 27, 28, 31, 37, 38, 43, 45, 46], "get": [3, 4, 5, 13, 16, 36, 47], "io": [3, 4, 6, 13, 16, 37, 40], "from": [3, 4, 6, 9, 10, 11, 13, 14, 15, 16, 17, 19, 20, 21, 22, 24, 25, 26, 27, 28, 32, 34, 37, 38, 39, 41, 43, 44, 45, 46, 47], "user": [3, 4, 10, 11, 17, 19, 20, 22, 26, 27, 35, 37, 38, 42, 43, 46, 47], "dummi": [3, 16, 43], "onnxmodelhandl": [3, 4, 11, 23, 37], "onnx_file_nam": 3, "inference_set": [3, 19, 37], "use_ort_extens": [3, 19], "external_initializers_file_nam": 3, "constant_inputs_file_nam": 3, "graph": [3, 4, 13, 20, 22, 25, 26, 37], "mixin": 3, "onnxepvalidatemixin": 3, "valid": [3, 4, 11, 22, 23, 37, 42, 43], "onnxgraphmixin": 3, "support": [3, 4, 9, 10, 11, 14, 15, 16, 21, 22, 23, 24, 26, 27, 29, 37, 38, 39, 42, 45, 46, 47], "oper": [3, 4, 5, 10, 13, 19, 20, 22, 23, 25, 26, 37], "distributedonnxmodelhandl": [3, 4], "openvinomodelhandl": [3, 4, 37], "main": [3, 4, 10, 31, 43], "respons": [3, 10, 13, 17], "snpemodelhandl": [3, 4, 37], "input_nam": [3, 4, 16, 20, 28, 37, 38, 39, 43], "input_shap": [3, 4, 16, 20, 28, 37, 38, 39, 43], "output_shap": [3, 28], "compositemodelhandl": [3, 4], "model_compon": 3, "model_component_nam": 3, "compositemodel": 3, "repres": [3, 11, 20, 22, 26, 27, 39, 47], "multipl": [3, 4, 8, 13, 18, 19, 20, 22, 29, 36, 44], "compon": [3, 4, 10, 22, 36], "onli": [3, 4, 8, 9, 10, 11, 14, 19, 21, 24, 26, 27, 28, 37, 41, 42, 43, 44, 47], "get_model_compon": 3, "iter": [3, 4, 10, 37], "child": 3, "whisper": [3, 4, 8, 22], "exampl": [3, 4, 14, 17, 32, 34, 36, 38, 39, 42, 43, 44, 46, 47], "ha": [3, 4, 10, 11, 14, 20, 22, 24, 31, 37, 39, 42, 45], "encod": [3, 4, 22], "decod": [3, 4, 22], "collect": [3, 9], "contain": [3, 4, 10, 14, 19, 23, 28, 37, 43, 46, 47], "should": [3, 4, 9, 11, 20, 37, 40, 41, 43], "have": [3, 4, 9, 10, 12, 14, 16, 19, 20, 22, 24, 37, 38, 44, 47], "same": [3, 4, 10, 11, 14, 16, 19, 20, 22, 26, 30, 37, 42, 47], "each": [4, 5, 6, 10, 11, 13, 19, 20, 22, 24, 37, 38, 39, 43, 47], "descript": [4, 5, 6, 11, 19, 20, 37, 38], "convert": [4, 13, 15, 18, 21, 22, 23, 25, 26, 27, 28, 37, 39, 43], "pytorch": [4, 8, 15, 18, 20, 25, 26, 29, 32, 37, 38, 39], "torch": [4, 13, 20, 21, 27, 38], "export": [4, 17, 19, 22, 30, 47], "cpu": [4, 7, 8, 11, 12, 13, 16, 19, 21, 24, 27, 31, 33, 37, 38, 47], "handler": 4, "hf": [4, 16, 17, 37, 43], "script": [4, 11, 19, 35, 37, 39, 40, 43], "other": [4, 5, 10, 11, 14, 16, 20, 22, 26, 38, 39, 40, 43, 46], "object": [4, 9, 10, 11, 22, 38], "import": [4, 9, 11, 16, 32, 34, 38, 39, 43, 44, 46, 47], "thi": [4, 5, 6, 9, 10, 11, 12, 13, 14, 16, 19, 20, 21, 22, 23, 24, 26, 27, 28, 30, 31, 32, 34, 35, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47], "pathlib": [4, 5], "default_valu": [4, 5, 6, 11, 26, 46], "searchable_valu": [4, 11, 26, 46], "depend": [4, 11, 13, 37, 39, 46, 47], "save_as_external_data": [4, 26], "serial": 4, "tensor": [4, 21, 22, 24, 27, 38], "data": [4, 9, 11, 19, 20, 26, 30, 35, 37, 44], "separ": [4, 9, 13, 17, 22, 31, 37], "instead": [4, 14, 20, 25, 28, 38, 47], "directli": [4, 9, 14, 37, 38, 43, 45], "larg": [4, 24, 37], "2gb": 4, "mai": [4, 14, 20, 22, 31, 38, 39, 40], "forc": [4, 13], "extern": [4, 22], "regardless": 4, "all_tensors_to_one_fil": [4, 26], "effect": 4, "If": [4, 5, 6, 11, 12, 13, 14, 16, 19, 20, 25, 26, 28, 30, 31, 37, 38, 41, 42, 43, 44, 45, 47], "specifi": [4, 9, 11, 12, 16, 19, 20, 24, 26, 37, 38, 40, 41, 42, 44, 45, 47], "external_data_nam": 4, "model_path_nam": 4, "size_threshold": 4, "threshold": [4, 37], "size": [4, 13, 14, 19, 20, 23, 24, 37, 41, 44], "when": [4, 6, 9, 10, 13, 14, 16, 19, 22, 23, 28, 37, 43, 46, 47], "To": [4, 10, 11, 12, 22, 30, 34, 35, 37, 38], "everi": [4, 36], "raw": [4, 43], "0": [4, 9, 12, 13, 16, 19, 22, 23, 24, 28, 37, 38, 39, 41, 42, 43, 44, 47], "1024": [4, 13, 22], "convert_attribut": 4, "non": [4, 22], "attribut": [4, 9, 16, 19, 47], "target_opset": [4, 20, 22, 37, 38, 39], "version": [4, 5, 7, 12, 13, 16, 19, 23, 31, 37, 45, 47], "default": [4, 5, 8, 10, 11, 14, 16, 19, 20, 24, 26, 30, 31, 37, 38, 39, 40, 41, 42, 43, 46, 47], "ai": [4, 8, 14, 17, 20, 23, 25, 28, 29, 30, 31, 37, 38, 47], "opset": [4, 20], "13": [4, 16, 20, 37, 38, 39], "use_dynamo_export": 4, "whether": [4, 11, 13, 14, 19, 22, 37, 43, 46], "dynamo_export": 4, "api": [4, 5, 13, 27, 36, 39], "past_key_value_nam": 4, "argument": [4, 37, 41, 43], "point": [4, 6, 10, 11, 22, 23, 25, 26, 28, 37], "For": [4, 11, 12, 13, 16, 17, 19, 20, 22, 37, 38, 39, 42, 43, 46, 47], "huggingfac": [4, 14, 24, 27, 35, 37, 38, 43], "past_key_valu": [4, 37], "basic": [4, 20, 22], "devic": [4, 10, 13, 22, 25, 37, 38, 42, 44, 47], "convers": [4, 8, 10, 18, 27, 32, 34, 35, 38, 39, 42], "e": [4, 13, 19, 22, 26, 31, 37, 42, 47], "g": [4, 13, 19, 22, 26, 42, 47], "cuda": [4, 8, 11, 13, 38, 42], "distributedhfmodel": 4, "torch_dtyp": [4, 13], "dtype": [4, 13, 17, 37], "cast": [4, 22], "befor": [4, 13, 14, 18, 26, 37, 40, 41, 43], "float32": [4, 13, 15, 17, 37], "float16": [4, 13, 15, 17, 37, 42], "parallel_job": 4, "number": [4, 5, 6, 13, 14, 19, 22, 27, 37, 41, 43], "parallel": [4, 13], "job": [4, 5, 16, 37, 40], "disabl": [4, 42], "merge_adapter_weight": 4, "merg": [4, 8, 22, 24, 37], "adapt": [4, 18, 24, 35, 37], "weight": [4, 8, 10, 13, 21, 22, 23, 24, 26, 27, 28, 37], "after": [4, 5, 13, 18, 22, 24, 32, 34, 37], "structur": [4, 6, 10, 21, 22, 24, 37], "consist": [4, 10], "base": [4, 11, 19, 22, 24, 27, 37, 42, 43, 47], "That": 4, "you": [4, 9, 12, 13, 14, 16, 19, 20, 22, 25, 27, 28, 30, 31, 32, 34, 35, 37, 38, 39, 40, 41, 43, 44, 45, 46, 47], "cannot": [4, 20, 22, 25, 28, 37, 38, 47], "some": [4, 5, 10, 11, 20, 22, 35, 37, 46, 47], "fine": [4, 8, 13, 17, 20, 22, 24], "tune": [4, 8, 10, 11, 17, 18, 24, 26, 27, 32, 36, 39], "save_metadata_for_token_gener": 4, "metadata": 4, "token": [4, 10, 13, 16, 24, 37, 43], "includ": [4, 11, 12, 13, 14, 19, 20, 22, 26, 30, 38, 43, 47], "generation_config": [4, 24], "relat": [4, 22, 37], "latest": [4, 22, 31], "21": 4, "optim": [4, 8, 10, 13, 16, 17, 20, 21, 23, 24, 27, 33, 35, 36, 37, 38, 39, 43, 47], "fuse": [4, 22, 26, 37], "node": [4, 13, 20, 22, 37, 41], "transform": [4, 8, 16, 17, 20, 21, 23, 24, 27, 32, 35, 37, 38, 43], "scenario": [4, 8, 19, 20, 22, 36, 37], "where": [4, 9, 12, 14, 17, 19, 20, 21, 22, 24, 37, 38, 41, 43, 47], "runtim": [4, 8, 10, 13, 19, 20, 22, 26, 28, 32, 37, 47], "doe": [4, 6, 22, 37, 42, 43, 47], "appli": [4, 16, 18, 19, 22, 24, 26, 37, 39], "time": [4, 14, 19, 20, 22, 23, 26, 30, 37, 41, 42, 47], "onnxruntim": [4, 11, 19, 20, 22, 29, 31, 37, 47], "model_typ": [4, 16, 22, 37], "bert": [4, 8, 16, 22, 33, 37, 42, 43], "gpt2": 4, "bert_tf": 4, "tf2onnx": 4, "bert_kera": 4, "keras2onnx": 4, "unet": 4, "vae": 4, "clip": 4, "stabl": [4, 6, 8, 31], "diffus": [4, 8], "num_head": [4, 37], "attent": [4, 37], "head": [4, 37], "num_key_value_head": 4, "hidden_s": [4, 37], "hidden": [4, 37], "optimization_opt": 4, "turn": [4, 10, 30], "off": [4, 24, 27, 30], "fusion": [4, 20], "fusion_opt": 4, "fusionopt": 4, "opt_level": [4, 42], "2": [4, 7, 13, 16, 17, 19, 21, 22, 24, 26, 27, 37, 38, 39, 41, 42, 43, 44, 47], "extend": 4, "99": [4, 19], "categor": [4, 11, 47], "use_gpu": 4, "flag": [4, 40, 41, 47], "gpu": [4, 7, 8, 11, 13, 20, 21, 22, 24, 27, 31, 42, 47], "infer": [4, 13, 20, 21, 22, 23, 24, 25, 26, 27, 28, 36, 37, 43, 44, 45, 47], "only_onnxruntim": 4, "python": [4, 5, 9, 10, 13, 19, 25, 28, 30, 31, 32, 34, 37, 38, 39, 41, 43, 44, 46], "might": [4, 10, 13, 26, 43], "caus": [4, 37], "failur": 4, "symbol": [4, 20], "shape": [4, 15, 22, 37, 38], "condit": [4, 11, 20], "parent": [4, 11, 37], "half": 4, "keep_io_typ": [4, 20], "keep": [4, 42], "origin": [4, 23, 24, 27, 42], "force_fp32_op": 4, "force_fp32_nod": 4, "force_fp16_input": 4, "even": [4, 27], "convert_float_to_float16": 4, "tool": [4, 22, 25, 26, 28, 35, 36, 37, 41, 42], "prefer": [4, 18, 20], "them": [4, 9, 16, 20, 22, 25, 26, 28, 41, 44], "use_gqa": 4, "replac": [4, 6, 10, 13, 37], "multiheadattent": 4, "groupqueryattent": 4, "input_int32": 4, "int32": [4, 22], "comput": [4, 8, 10, 16, 19, 22, 23, 26, 28, 37, 40, 41, 47], "select": [4, 10, 26, 34, 38, 40], "process": [4, 8, 9, 13, 14, 16, 18, 23, 24, 28, 30, 37, 38, 39, 43], "cpu_cor": [4, 13], "core": [4, 10, 13, 14, 25, 28, 36], "thread": [4, 13, 22], "io_bind": [4, 13, 19, 42], "enabl": [4, 13, 14, 16, 27, 31, 36, 37, 42], "iobind": [4, 13], "enable_cuda_graph": [4, 13, 42], "providers_list": [4, 13, 22], "cpuexecutionprovid": [4, 11, 19, 22, 47], "execution_mode_list": [4, 13], "between": [4, 5, 10, 12, 13, 16, 22, 27, 37, 42], "opt_level_list": [4, 13], "trt_fp16_enabl": [4, 13], "fp16": [4, 8, 13, 21, 27, 42], "mode": [4, 13, 22, 31, 37, 38], "tensorrt": [4, 11, 13, 21, 24], "intra_thread_num_list": [4, 13], "intra": [4, 13], "test": [4, 13, 28, 31, 37], "inter_thread_num_list": [4, 13], "inter": [4, 13], "extra_session_config": [4, 13, 19], "extra": [4, 13, 19, 31, 47], "session": [4, 22, 30], "dure": [4, 13, 22, 23, 24, 31, 37], "force_evaluate_other_ep": 4, "differ": [4, 9, 10, 13, 19, 22, 24, 25, 28, 37, 38, 42, 44, 47], "associ": [4, 13], "enable_profil": [4, 13, 22, 37], "profil": [4, 13], "min_positive_v": 4, "constant": [4, 22, 35], "against": [4, 47], "1e": 4, "07": 4, "max_finite_v": 4, "10000": 4, "left": 4, "use_symbolic_shape_inf": 4, "op_block_list": [4, 20], "op": [4, 18, 20, 37], "leav": [4, 20], "node_block_list": 4, "name_pattern": 4, "whose": 4, "match": 4, "pattern": [4, 20, 21, 24], "By": [4, 37], "defaultlook": 4, "logit": [4, 16, 20, 44], "mix": [4, 15, 37], "simplifiedlayernorm": [4, 20], "skipsimplifiedlayernorm": [4, 20], "relu": [4, 20], "add": [4, 14, 16, 19, 20, 30, 35, 43, 45], "atol": 4, "absolut": [4, 23], "toler": 4, "check": [4, 14, 22, 24, 26, 27, 28, 30, 46], "06": 4, "preprocess": [4, 26, 43], "quantiz": [4, 8, 10, 18, 27, 32, 34, 35, 37, 38, 39], "qnn": [4, 8, 13, 18, 20, 26, 37], "fuse_layernorm": 4, "reducemean": 4, "sequenc": [4, 13, 37], "singl": [4, 8, 13, 19, 22, 26, 27, 36, 43], "layernorm": [4, 20], "inputs_to_make_channel_last": 4, "transpos": 4, "channel": 4, "last": [4, 38, 40], "input0": 4, "n": [4, 13, 17], "c": [4, 13, 24, 25, 26, 37, 41], "d1": 4, "d2": 4, "dn": 4, "chang": [4, 22, 38], "updat": [4, 20, 35, 38, 39], "input0_chanfirst": 4, "potenti": [4, 14], "improv": [4, 18, 20, 22, 23, 24, 26, 27, 35, 37, 42, 44], "qdq": [4, 8, 26], "ep": [4, 8, 10, 26, 42, 47], "becaus": 4, "addit": [4, 17, 23, 26, 31, 38, 43, 45], "allow": [4, 12, 13, 16, 20, 22, 23, 30, 36, 43], "insert": [4, 18, 23, 26], "ort": [4, 17, 18, 26], "layout": [4, 22], "cancel": 4, "out": [4, 22, 24, 26, 27, 28, 40, 46, 47], "outputs_to_make_channel_last": 4, "output0": 4, "output0_chanfirst": 4, "overrid": [4, 16, 19, 37, 39], "pre": [4, 8, 18, 24, 26, 35, 37, 46, 47], "resolv": 4, "constraint": [4, 36], "being": [4, 47], "refer": [4, 5, 6, 12, 16, 20, 21, 22, 23, 24, 25, 26, 27, 28, 31, 36, 37, 38, 39, 43, 44, 47], "situat": 4, "certain": [4, 10, 16, 20, 46], "16": [4, 13, 17, 22, 24, 44], "bit": [4, 8, 23, 24, 25, 26, 27, 28], "standalon": 4, "rather": 4, "neighbor": 4, "well": [4, 17, 23, 36, 43], "specif": [4, 9, 10, 16, 18, 19, 20, 22, 25, 26, 36, 37, 39, 46], "problem": 4, "aris": 4, "here": [4, 9, 12, 14, 16, 20, 22, 37, 38, 40, 42, 43, 44, 47], "note": [4, 11, 12, 16, 19, 20, 23, 24, 25, 26, 28, 31, 37, 38, 39, 40, 43, 45, 47], "handl": [4, 39], "just": [4, 38, 43], "initi": [4, 5, 10, 13, 22, 24, 37, 39], "activ": [4, 10, 23, 24, 26, 27, 31, 42], "overrides_config": 4, "format": [4, 13, 18, 19, 20, 22, 25, 26, 37], "tensor_nam": 4, "quant_typ": [4, 24], "element_wise_binary_op": 4, "element": 4, "wise": 4, "binari": [4, 18, 37], "sub": [4, 10], "mul": 4, "div": 4, "dynam": [4, 8, 15, 26, 37, 38], "quant_mod": 4, "weight_typ": [4, 26, 37, 39, 46], "both": [4, 20, 26, 39, 42, 44, 45, 47], "static": [4, 8, 11, 26, 37], "qint8": [4, 26], "sign": 4, "8": [4, 7, 13, 16, 23, 24, 25, 26, 27, 28, 43, 47], "integ": [4, 22, 23], "quint8": [4, 26, 37, 39, 46], "unsign": 4, "op_types_to_quant": 4, "append_first_op_types_to_quantize_list": 4, "append": [4, 9, 18], "firstli": [4, 19], "appear": 4, "nodes_to_quant": 4, "nodes_to_exclud": 4, "exclud": [4, 22, 37], "per_channel": [4, 26, 46], "per": [4, 10], "tip": 4, "reduce_rang": [4, 46], "http": [4, 5, 6, 13, 14, 31, 37, 43], "doc": [4, 42], "perform": [4, 10, 13, 18, 19, 20, 24, 26, 27, 32, 37, 38], "html": [4, 6, 13], "reduc": [4, 22, 23, 24, 26, 27, 36], "rang": [4, 22], "7": 4, "vnni": 4, "machin": [4, 10, 16, 19, 20, 22, 26, 41, 45, 47], "especi": 4, "quant_preprocess": 4, "prepar": [4, 14, 16, 36], "sigmoid": 4, "nnapi": [4, 20], "activationsymmetr": 4, "symmetr": [4, 13, 17], "calibr": [4, 25, 26], "weightsymmetr": 4, "enablesubgraph": 4, "subgraph": 4, "current": [4, 10, 13, 14, 35, 37, 42, 43, 45, 47], "forcequantizenoinputcheck": 4, "latent": 4, "like": [4, 5, 14, 19, 21, 22, 24, 26, 37, 38, 41, 43], "maxpool": 4, "do": [4, 11, 16], "alreadi": [4, 17, 19], "alwai": [4, 37], "behavior": [4, 5], "matmulconstbonli": 4, "matmul": 4, "const": 4, "b": [4, 20, 22, 24, 26], "conditionaldefault": [4, 11], "olive_invalid_param_valu": 4, "extra_opt": 4, "pair": 4, "pleas": [4, 5, 11, 12, 14, 16, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 31, 32, 34, 37, 38, 40, 44, 47], "github": [4, 13, 31, 37, 43], "com": [4, 5, 12, 31, 37, 43, 47], "microsoft": [4, 5, 12, 13, 31, 37, 40, 43, 47], "blob": [4, 43], "py": [4, 9, 19, 22, 24, 26, 28, 34, 37, 43, 44], "detail": [4, 5, 6, 12, 16, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 36, 37, 38, 39, 40, 41, 47], "about": [4, 6, 11, 14, 16, 20, 22, 23, 24, 25, 26, 27, 28, 37, 38, 39], "overwritten": 4, "calibrate_method": [4, 26], "method": [4, 11, 13, 17, 23, 24, 26, 37], "minmax": 4, "entropi": 4, "calibrationmethod": 4, "percentil": 4, "avoid": [4, 38, 43], "quant_format": [4, 26], "qoper": 4, "quantizelinear": 4, "dequantizelinear": 4, "activation_typ": [4, 26], "specialparamvalu": 4, "invalid": [4, 11, 22], "prepare_qnn_config": 4, "suitabl": [4, 20, 36], "qnn_extra_opt": 4, "execution_provid": [4, 7, 11, 19, 37, 42, 44, 47], "get_qnn_qdq_config": 4, "init_overrid": 4, "copi": 4, "necessari": [4, 14, 16, 26, 30, 32, 34, 38, 47], "adjust": 4, "tensorquantoverrid": 4, "either": [4, 9, 26, 37, 47], "assum": [4, 16, 37], "axi": [4, 37], "must": [4, 10, 11, 16, 22, 25, 26, 37, 43, 46, 47], "present": [4, 37], "first": [4, 10, 16, 19, 32, 34, 37, 39, 43], "quanttyp": 4, "scale": [4, 13, 26, 41], "zero_point": 4, "zero": [4, 26], "rmax": 4, "maximum": [4, 13, 17, 37, 41], "real": 4, "rmin": 4, "minimum": [4, 13, 37, 41], "A": [4, 10, 19, 25, 41, 47], "nest": 4, "anoth": [4, 11, 15, 18, 20, 37, 38, 42, 43], "recv_nod": 4, "consum": [4, 13], "add_qtype_convert": 4, "automat": [4, 9, 12, 13, 16, 22, 30, 37, 38, 42, 45], "entri": 4, "ensur": 4, "ex": 4, "up": [4, 24, 27, 36, 42], "surround": 4, "appropri": 4, "document": [4, 12, 16, 22, 28, 36, 37, 40, 44, 45, 46], "mixed_precision_overrides_util": 4, "further": 4, "techniqu": [4, 10, 11, 22, 24, 26, 27, 36, 37, 38, 39], "ignor": [4, 5, 11, 13, 19, 41, 43], "olive_ignored_param_valu": 4, "fix": [4, 10, 15, 25, 26, 28, 37, 46], "dim_param": 4, "dim_valu": [4, 20], "comma": [4, 13], "3": [4, 5, 13, 19, 20, 23, 24, 25, 26, 27, 28, 37, 38, 39, 43], "256": [4, 13, 17], "intel": [4, 8, 23, 29, 37, 42, 43], "neural": [4, 8, 23, 28, 29, 37], "compressor": [4, 8, 29, 37], "approach": [4, 23, 24, 26, 39], "onnxrt_trt_ep": 4, "onnxrt_cuda_ep": 4, "domain": 4, "auto": [4, 10, 11, 35, 39], "cv": 4, "object_detect": 4, "nlp": [4, 8], "recommendation_system": 4, "adaptor": 4, "explicitli": [4, 45], "detect": 4, "workspac": [4, 5, 10, 13, 16, 19, 37, 40, 41, 45, 47], "intermedi": [4, 10, 14], "histori": [4, 19], "store": [4, 14, 30, 37, 38, 45], "nc_workspac": 4, "datetim": 4, "now": [4, 23, 37, 38, 39, 42, 43], "strftime": 4, "y": 4, "m": [4, 10, 13, 17, 25, 28, 30, 32, 34, 38], "d_": 4, "h": [4, 13, 17], "recip": 4, "below": [4, 47], "smooth_quant": 4, "smooth": 4, "quant": 4, "smooth_quant_arg": 4, "fast_bias_correct": 4, "fast": 4, "bia": 4, "correct": 4, "weight_correct": 4, "gemm_to_matmul": 4, "gemm": 4, "graph_optimization_level": [4, 19], "disable_al": 4, "enable_bas": 4, "enable_extend": 4, "enable_al": 4, "first_conv_or_matmul_quant": 4, "conv": 4, "last_conv_or_matmul_quant": 4, "pre_post_process_quant": 4, "postprocess": 4, "add_qdq_pair_to_weight": 4, "optypes_to_exclude_output_qu": 4, "don": [4, 11, 12, 20, 22, 37], "t": [4, 11, 12, 19, 20, 22, 37, 45, 47], "optyp": 4, "dedicated_qdq_pair": 4, "dedic": 4, "quant_level": 4, "choos": [4, 22], "conserv": 4, "combin": [4, 10, 27, 42, 46], "master": [4, 24], "tuning_strategi": 4, "md": 4, "algorithm": [4, 6, 24, 26, 27, 29, 37], "excluded_precis": 4, "empti": [4, 37], "fp32": [4, 8, 42], "bf16": 4, "int8": [4, 8, 26, 42], "want": [4, 14, 16, 19, 20, 30, 31, 37, 38, 39, 43], "tuning_criterion": 4, "instanc": [4, 5, 13, 19, 22, 39, 41, 47], "tuningcriterion": 4, "In": [4, 10, 12, 17, 20, 22, 26, 37, 39, 40, 42, 43, 44, 47], "strategy_kwarg": 4, "timeout": [4, 5, 37], "max_trial": 4, "5": [4, 5, 11, 19, 24, 34, 37, 39], "awar": [4, 18, 27, 36, 37, 42], "weight_only_config": [4, 26], "inc": [4, 25], "op_type_dict": 4, "4": [4, 7, 8, 9, 11, 19, 21, 22, 24, 26, 27, 37, 43], "group_siz": 4, "scheme": [4, 24], "asym": 4, "rtn": 4, "calibration_sampling_s": [4, 26], "sampl": [4, 6, 10, 13, 37], "100": [4, 42], "weight_onli": [4, 26], "we": [4, 10, 22, 28, 31, 36, 37, 38, 39, 40, 42, 43, 46], "best": [4, 10, 19, 20, 22, 32, 34, 36, 37, 39, 42, 44, 46], "vai_q_onnx": 4, "viti": [4, 8, 29, 37], "input_nod": 4, "start": [4, 19, 36, 39, 43], "need": [4, 10, 12, 14, 16, 18, 22, 26, 28, 30, 31, 37, 39, 42, 43, 44], "output_nod": 4, "end": [4, 37], "optimize_model": 4, "deprec": 4, "soon": [4, 39], "NOT": [4, 42], "recommend": [4, 28, 31], "make": [4, 14, 16, 20, 24, 26, 27, 36, 40, 42], "debug": 4, "loss": 4, "difficult": 4, "use_external_data_format": 4, "nonoverflow": [4, 26], "minms": 4, "need_layer_fus": 4, "layer": [4, 21, 24, 37], "enable_dpu": 4, "dpu": [4, 8], "addqdqpairtoweight": 4, "remain": 4, "post": [4, 8, 9, 18, 24, 37, 43], "command": [4, 14, 16, 17, 22, 30, 31, 32, 34, 35, 38, 39, 40, 41], "tool_command": [4, 22], "composit": 4, "invok": [4, 11, 13], "tool_command_arg": [4, 22], "prepostprocessor": [4, 22], "schema": [4, 37], "would": [4, 38, 42, 43], "imag": [4, 19, 22, 47], "data_typ": [4, 22], "uint8": [4, 22], "num_byt": [4, 22], "append_pre_post_processing_op": 4, "prepostprocessorinput": [4, 22], "beam": [4, 18], "whisperbeamsearch": 4, "contrib": 4, "17": 4, "els": 4, "beamsearch": 4, "no_repeat_ngram_s": [4, 22], "ngram": 4, "occur": 4, "onc": [4, 14, 23, 26], "use_vocab_mask": 4, "vocab_mask": 4, "use_prefix_vocab_mask": 4, "prefix_vocab_mask": 4, "use_forced_decoder_id": 4, "decoder_input_id": 4, "use_logits_processor": 4, "logits_processor": 4, "use_temperatur": 4, "temperatur": 4, "extract": [4, 18, 37], "make_input": [4, 22], "proto": 4, "exist": [4, 13, 14, 16, 19, 20, 22, 41, 47], "sessionopt": 4, "add_external_initi": [4, 22], "pack_input": [4, 22], "pack": [4, 13, 22, 27], "modul": [4, 8, 9, 13, 20, 21, 37, 43], "dynamic_lora_r": 4, "lora_r": [4, 13], "optional_input": 4, "dimens": [4, 20, 22, 37], "save_format": [4, 13], "common": [4, 7, 47], "util": [4, 36, 37], "weightsfileformat": 4, "numpi": [4, 13], "hug": [4, 21, 24], "face": [4, 21, 24], "target_modul": [4, 13], "r": [4, 13, 32, 34], "64": [4, 13, 17, 19], "32": [4, 13, 17, 23, 26, 38, 39], "lora_alpha": [4, 13, 24], "alpha": [4, 13], "lora_dropout": 4, "dropout": 4, "probabl": 4, "05": 4, "modules_to_sav": 4, "apart": 4, "trainabl": [4, 22, 24], "checkpoint": [4, 13], "train": [4, 8, 13, 16, 17, 18, 20, 22, 36, 37], "bfloat16": [4, 13, 24], "allow_tf32": 4, "tf32": 4, "amper": 4, "speed": [4, 14, 26, 27], "inform": [4, 6, 11, 12, 23, 24, 39, 42], "see": [4, 20, 22, 37, 38, 39, 40], "org": 4, "tensorfloat": 4, "train_data_config": [4, 24], "eval_data_config": 4, "training_arg": [4, 24], "double_qu": 4, "again": 4, "fp4": 4, "nf4": [4, 24], "compute_dtyp": [4, 24], "loftq_it": 4, "trainer": 4, "trainingargu": 4, "paged_adamw_32bit": 4, "learning_r": [4, 24], "0002": [4, 24], "learn": [4, 5, 14, 16, 19, 20, 21, 22, 23, 24, 26, 27, 28, 36, 41, 45, 47], "rate": 4, "adamw": 4, "gradient_checkpoint": 4, "gradient": [4, 24], "lr_scheduler_typ": 4, "schedul": 4, "better": [4, 37, 42], "than": [4, 26, 38, 39], "cosin": 4, "advantag": [4, 21, 39], "analysi": 4, "warmup_ratio": 4, "03": 4, "fraction": 4, "warmup": 4, "evaluation_strategi": 4, "eval_dataset": 4, "otherwis": [4, 13, 37, 47], "unless": 4, "epoch": 4, "report_to": 4, "integr": [4, 8, 20, 22, 26, 27, 35, 36], "report": 4, "dir": 4, "temp": 4, "overwrite_output_dir": 4, "overwrit": 4, "content": [4, 25], "continu": [4, 14, 30], "resume_from_checkpoint": 4, "folder": [4, 5, 11, 12, 13, 19, 37], "superced": 4, "found": [4, 14, 16, 20, 22, 24, 25, 28, 37, 38, 39, 41], "extra_arg": [4, 25], "keyword": [4, 37, 43], "create_training_arg": 4, "val_data_config": 4, "training_loop_func": [4, 24], "loop": [4, 8, 24], "ptl_modul": [4, 24], "lightningmodul": 4, "lightn": [4, 8, 24], "wai": [4, 15, 22, 24, 26, 43, 44], "encapsul": 4, "logic": [4, 9], "readthedoc": [4, 6], "en": [4, 5, 6], "lightning_modul": 4, "ptl_data_modul": [4, 24], "lightningdatamodul": 4, "datamodul": 4, "num_epoch": [4, 24, 34], "num_step": 4, "do_valid": 4, "over": [4, 6, 10, 37, 39], "modules_to_fus": 4, "qconfig_func": 4, "qconfig": [4, 24], "qat": [4, 8, 24], "ao": 4, "pytorch_lightn": 4, "seed": [4, 6, 19, 37, 42], "random": [4, 6, 10, 37], "checkpoint_path": 4, "tensorflow": [4, 13, 25, 26, 28, 37], "openvino": [4, 8, 18, 37, 38, 42, 47], "your": [4, 8, 9, 13, 14, 16, 19, 22, 25, 27, 28, 30, 31, 37, 38, 39, 40, 41, 43, 44, 45], "example_input_func": 4, "tf": 4, "ndarrai": 4, "compress_to_fp16": 4, "compress": [4, 26, 36], "extra_config": [4, 19], "2023": [4, 23], "openvino_docs_ov_converter_ug_conversion_opt": 4, "output_model": 4, "ov_model": 4, "special": [4, 17, 38], "preserv": 4, "distilbert": 4, "etc": [4, 11, 17, 22, 26, 37, 38, 42, 47], "modeltypeenum": 4, "preset": 4, "defin": [4, 9, 10, 12, 13, 19, 26, 43, 44, 46], "presetenum": 4, "ignored_scop": 4, "basic_quantization_flow": 4, "ignored_scope_typ": 4, "scope": 4, "ignorescopetypeenum": 4, "target_devic": 4, "cpu_spr": [4, 7], "vpu": [4, 7], "workflow": [4, 9, 10, 12, 13, 14, 19, 32, 34, 35, 38, 39, 45], "hardwar": [4, 7, 8, 10, 18, 20, 22, 23, 26, 27, 36, 37, 38, 47], "snpe": [4, 8, 13, 18, 20, 37, 38, 42], "dlc": [4, 28, 37], "sdk": [4, 18, 37], "tensorflowmodelhandl": [4, 23], "length": [4, 13, 20, 37], "input_typ": [4, 16, 37, 43], "member": [4, 11], "platform_sdk": 4, "qualcomm": [4, 8, 25, 28], "inputtyp": 4, "input_layout": 4, "inputlayout": 4, "develop": [4, 25, 26, 28, 31, 36, 42], "site": 4, "string": [4, 5, 11, 13, 16, 22, 37], "enable_cpu_fallback": 4, "priority_hint": 4, "low": [4, 24], "use_enhanced_quant": 4, "enhanc": 4, "featur": [4, 13, 14, 31, 38, 39], "determin": [4, 42, 44], "long": [4, 30, 37], "tail": 4, "distribut": [4, 23, 37], "enable_htp": [4, 28], "htp": 4, "window": [4, 14, 38], "htp_soc": 4, "soc": 4, "offlin": [4, 22], "tools_snp": 4, "bias_bitwidth": 4, "overwrite_cache_record": 4, "oliveol": 4, "snpedevic": 4, "12": [4, 7, 37], "input_list": [4, 25], "qnnmodelhandl": 4, "input_dim": [4, 20], "network": [4, 23, 24, 26, 28, 37], "224": [4, 20, 43], "quot": 4, "charact": 4, "space": [4, 6, 10, 11, 37, 38, 46], "line": [4, 16, 17, 22, 35, 37, 38, 39, 41], "data2": 4, "out_nod": 4, "out_1": 4, "out_2": 4, "show_unconsumed_nod": 4, "custom_io": 4, "bundl": 4, "publicresourc": 4, "topic": 4, "80": 4, "63442": 4, "50": [4, 17], "compil": [4, 21, 25, 36], "code": [4, 13, 17, 19, 24, 31, 32, 34, 38, 43], "librari": [4, 18, 20, 21, 22, 24, 26, 27, 37], "lib": 4, "lib_target": [4, 25], "build": [4, 10, 19, 31, 36, 47], "aarch64": 4, "android": 4, "x86_64": [4, 25], "linux": [4, 25, 38], "clang": [4, 25], "lib_nam": 4, "bin": [4, 47], "qnn_model": 4, "context": [4, 18, 24, 37, 47], "particular": 4, "binary_fil": 4, "extens": [4, 21, 22, 36, 37], "arxiv": 4, "ab": 4, "2301": 4, "00774": 4, "bloom": [4, 24], "gpt_neox": 4, "llama": [4, 8, 16, 17, 37, 43], "opt": [4, 24], "sparsiti": [4, 8, 21, 24], "two": [4, 10, 22, 26, 46, 47], "semi": [4, 24], "nvidia": 4, "blog": [4, 24], "blocksiz": 4, "mask": 4, "128": [4, 13, 17, 43], "percdamp": 4, "percentag": [4, 37], "averag": 4, "hessian": 4, "diagon": 4, "dampen": 4, "01": [4, 9, 23, 37, 42, 44], "min_lay": 4, "prune": [4, 24, 26], "id": [4, 5, 12, 13, 19, 38, 41, 45, 47], "max_lay": 4, "layer_name_filt": 4, "given": [4, 13, 36, 37, 42, 43], "affect": [4, 14], "expect": [4, 24, 42], "most": [4, 20, 22, 27, 36, 37, 42, 47], "pdf": 4, "2401": 4, "15024": 4, "calibration_data_config": [4, 24], "dataset": [4, 9, 17, 24, 25, 26, 38, 43], "calcul": [4, 26, 44], "calibration_nsampl": 4, "calibration_batch_s": 4, "batch": [4, 13, 19, 20, 37, 43, 44], "42": 4, "measur": [4, 37], "how": [4, 12, 16, 22, 35, 36, 37, 38, 39, 40], "much": 4, "slice": [4, 24], "round_interv": 4, "interv": [4, 5, 37], "round": 4, "final_orient": 4, "orient": 4, "choic": [4, 13], "pca": 4, "new": [4, 13, 16, 24, 35, 40, 41, 47], "rotat": [4, 27], "abl": [4, 37, 40], "llm": [4, 27], "2404": 4, "00456": 4, "input_model_dtyp": 4, "modeldtyp": 4, "hadamard": 4, "rotation_se": 4, "matrix": 4, "replic": 4, "paper": [4, 24, 27], "w_rtn": [4, 27], "w_gptq": 4, "gptq": [4, 26, 27, 37], "flash_attention_2": 4, "newer": 4, "gptq_damp": 4, "damp": 4, "factor": [4, 14, 42], "gptq_opt_scal": 4, "w_bit": [4, 27], "w_asym": 4, "asymmetr": [4, 13, 17], "w_groupsiz": 4, "group": [4, 5, 6, 13, 37, 41, 47], "groupwis": 4, "a_bit": [4, 27], "a_asym": 4, "a_clip_ratio": 4, "ratio": 4, "new_max": 4, "clip_ratio": 4, "a_groups": 4, "k_bit": [4, 27], "k_clip_ratio": 4, "k_groupsiz": 4, "v_bit": [4, 27], "v_clip_ratio": 4, "v_groupsiz": 4, "optimum": [4, 8, 16, 37], "kernel": 4, "nsampl": 4, "layers_block_nam": 4, "block": [4, 10, 13], "fill": [4, 40, 43], "link": [4, 8, 12, 24], "autogptq": [4, 29], "896d8204bc89a7cfbda42bf3314e13cf4ce20b02": 4, "auto_gptq": 4, "l19": 4, "l26": 4, "outside_layer_modul": 4, "nn": [4, 21], "inside_layer_modul": 4, "linear": [4, 21], "damp_perc": 4, "static_group": 4, "true_sequenti": 4, "sequenti": 4, "desc_act": 4, "sym": 4, "awq": [4, 27, 37], "autoawq": [4, 29], "q_group_siz": 4, "column": [4, 24], "big": [4, 27], "batch_siz": [4, 9, 13, 16, 20, 22, 37, 39, 42, 43, 44], "gemv": 4, "compat": [4, 22, 26], "exllama": 4, "duo_scal": 4, "w": 4, "x": 4, "modules_to_not_convert": 4, "llava": 4, "mixtral": 4, "gate": [4, 16], "export_compat": 4, "down": [4, 30, 41], "pack_model_for_onnx_convers": [4, 27], "spars": [4, 21, 24], "applic": [4, 21], "entir": [4, 20, 37], "instal": [4, 10, 11, 16, 19, 32, 34, 35, 38, 40, 47], "dataload": [4, 9, 43], "14": [4, 19, 20], "decoder_model": 4, "decoder_with_past_model": 4, "mean": [4, 16, 26, 37], "main_export": 4, "its": [4, 5, 6, 10, 16, 20, 22, 23, 24, 25, 26, 27, 28, 37, 44, 47], "via": [4, 12, 37], "strict": 4, "decoder_with_past": 4, "strictli": 4, "case": [4, 13, 20, 22, 25, 32, 34, 37, 42, 43, 46, 47], "ad": [4, 9, 16, 22, 37], "builder": 4, "genai": 4, "model_build": 4, "metadata_onli": 4, "int4_block_s": [4, 13, 17], "block_siz": 4, "int4": [4, 8, 20, 42], "accept": [4, 11, 38], "int4_accuracy_level": 4, "accuracylevel": 4, "exclude_emb": 4, "remov": [4, 13], "embed": 4, "exclude_lm_head": 4, "languag": [4, 17, 27, 37], "captur": 4, "place": [4, 9], "prerequisit": [4, 18], "correctli": 4, "resourc": [5, 7, 12, 13, 16, 19, 37, 41, 45, 47], "resource_path": 5, "azureml_datastor": [5, 12, 37, 45], "azureml_job_output": [5, 12, 45], "azureml_model": [5, 12, 45], "azuremlregistrymodel": 5, "azureml_registry_model": [5, 12, 16, 45], "string_nam": [5, 45], "local": [5, 7, 8, 10, 12, 13, 30, 37, 38, 39], "azureml": [5, 7, 8, 10, 13, 14, 19, 40, 41], "azureml_cli": [5, 12, 16, 45, 47], "client": [5, 47], "azuremlclientconfig": [5, 7], "azuremlcli": 5, "mlclient": [5, 45], "read_timeout": [5, 37], "max_operation_retri": [5, 37], "operation_retry_interv": [5, 37], "control": [5, 20], "creation": [5, 19, 37], "download": [5, 12, 14, 16, 24, 37, 40, 47], "subscription_id": [5, 12, 13, 16, 37, 41, 45, 47], "azur": [5, 13, 19, 35, 45, 47], "subscript": [5, 13, 37, 41, 47], "aml_config_path": [5, 13, 41], "resource_group": [5, 12, 13, 16, 37, 41, 45, 47], "workspace_nam": [5, 12, 13, 16, 37, 41, 45, 47], "60": [5, 24], "read": [5, 23, 37], "second": [5, 10, 13, 37, 41], "request": [5, 37], "retri": [5, 37], "doubl": [5, 37], "default_auth_param": [5, 37], "auth": [5, 37], "u": [5, 22, 35, 39], "ident": [5, 16, 37, 47], "defaultazurecredenti": [5, 37], "view": 5, "keyvault_nam": [5, 13, 16, 37], "keyvault": [5, 13, 16, 37], "retriev": [5, 16, 37], "secret": [5, 16, 37], "get_workspace_config": 5, "create_cli": 5, "create_registry_cli": 5, "registry_nam": [5, 12, 13, 16, 45], "datastor": [5, 7, 30, 37, 47], "datastore_nam": [5, 12, 37, 45], "relative_path": [5, 12, 37, 45], "rel": [5, 12, 45], "datastore_url": 5, "url": [5, 14, 37], "job_nam": 5, "exhaust": [6, 10, 22, 37, 39], "grid": 6, "searcher": 6, "without": [6, 10, 14, 27, 30, 37], "num_sampl": [6, 19, 37, 42], "suggest": [6, 35], "rng": 6, "with_replac": 6, "tpe": [6, 10, 19, 37, 42], "tree": [6, 10, 37], "parzen": [6, 10, 37], "estim": [6, 10, 37], "optuna": 6, "tpesampl": 6, "underneath": 6, "sampler": 6, "multivari": 6, "decompos": 6, "docker": [7, 10, 19, 37, 40], "isolatedort": [7, 47], "pythonenviron": [7, 37, 47], "intel_myriad": 7, "npu": [7, 8, 25, 47], "hf_token": [7, 16, 37], "local_docker_config": [7, 47], "localdockerconfig": [7, 47], "is_dev": 7, "requirements_fil": [7, 19, 47], "image_nam": [7, 47], "dockerfil": [7, 47], "build_context_path": [7, 47], "build_arg": 7, "run_param": 7, "aml_comput": [7, 12, 13, 37, 47], "aml_docker_config": [7, 12, 37, 47], "azuremldockerconfig": 7, "aml_environment_config": [7, 47], "azuremlenvironmentconfig": 7, "tag": [7, 47], "workspaceblobstor": [7, 30, 47], "instance_count": [7, 19, 47], "base_imag": [7, 12, 37, 47], "conda_file_path": [7, 12, 37, 47], "python_environment_path": [7, 47], "environment_vari": [7, 19, 47], "prepend_to_path": [7, 47], "kwarg": [7, 37, 43], "python_environ": [7, 47], "isolated_ort": [7, 47], "system_alia": 7, "azurend12ssystem": 7, "classvar": 7, "num_cpu": 7, "num_gpu": 7, "sku": [7, 19], "standard_nd12": 7, "system_typ": 7, "azurend24rssystem": 7, "24": 7, "standard_nd24r": 7, "azurend24ssystem": 7, "standard_nd24": 7, "azurend6ssystem": 7, "6": [7, 11, 13, 28], "standard_nd6": 7, "azurend96a100system": 7, "96": 7, "standard_nd96amsr_a100_v4": 7, "azurend96asystem": 7, "standard_nd96asr_v4": 7, "azurendv2system": [7, 37, 47], "40": 7, "standard_nd40rs_v2": 7, "surfaceprosystem1796": 7, "surface_pro_1796": 7, "surfaceprosystem1807": 7, "surface_pro_1807": 7, "llama2": 8, "qlora": [8, 13, 17, 18, 22], "mistral": 8, "open": [8, 20, 22, 23, 26], "sparsegpt": [8, 18], "torchtrt": 8, "phi2": 8, "slicegpt": [8, 18], "falcon": 8, "red": 8, "pajama": 8, "trt": [8, 42], "deberta": 8, "registri": [8, 9, 13, 43], "gptj": 8, "audio": 8, "spectrogram": 8, "vision": 8, "xl": 8, "directml": [8, 11, 31, 47], "toolkit": [8, 23], "squeezenet": 8, "mobilenet": 8, "resnet": [8, 33, 37, 38, 39], "amd": [8, 29, 37], "vgg": 8, "incept": [8, 43], "super": 8, "resolut": 8, "abil": [9, 26], "write": [9, 22, 23, 39, 42], "own": [9, 10, 13, 16, 22, 30, 36, 37, 43, 44, 45], "evaluate_func": [9, 37, 44], "metric_func": [9, 37, 44], "openvinoquant": [9, 23, 37], "my_script": 9, "post_process": [9, 37, 43, 44], "mydataload": 9, "def": [9, 11, 43, 44], "__init__": 9, "self": [9, 11, 12, 35], "__len__": 9, "__getitem__": 9, "register_dataload": [9, 43], "my_dataload": 9, "register_post_process": [9, 43], "my_post_process": 9, "accuracy_data_config": [9, 37, 44], "load_dataset_config": [9, 16, 42, 43], "skip_dataset": 9, "post_process_data_config": [9, 37, 43, 44], "dataloader_config": [9, 22, 24, 26, 28, 37, 42, 43, 44], "param": [9, 11, 22, 37, 43, 44], "prioriti": [9, 16, 19, 37, 39, 42, 44], "goal": [9, 10, 26, 37, 42, 44], "degrad": [9, 37, 42, 44], "sy": 9, "interpret": 9, "let": [9, 38, 39, 43, 46], "sai": [9, 10, 11], "my_modul": 9, "my_olive_project": 9, "print": 9, "my_val": 9, "section": [10, 12, 16, 37, 43, 45], "discuss": 10, "concept": [10, 47], "compos": [10, 36, 37, 43], "modular": 10, "construct": 10, "achiev": [10, 26], "respect": [10, 26], "emploi": 10, "togeth": [10, 22, 37], "desir": [10, 25, 26, 31, 36, 38, 39], "pareto": [10, 37], "frontier": [10, 37], "multi": 10, "meet": [10, 43], "diagram": 10, "illustr": 10, "relationship": 10, "searchabl": [10, 11, 26, 37, 46], "chose": [10, 39, 46], "possibl": [10, 13, 20, 27, 38, 46, 47], "environ": [10, 13, 19, 20, 22, 25, 26, 28, 30, 31, 37, 38, 40], "o": [10, 13, 17, 38, 47], "platform": [10, 20, 22, 23, 26, 47], "thu": [10, 22, 26, 37], "ideal": 10, "deploy": [10, 19, 36], "five": [10, 47], "localsystem": [10, 37, 39, 42, 47], "azuremlsystem": 10, "dockersystem": [10, 47], "pythonenvironmentsystem": [10, 47], "isolatedortsystem": [10, 47], "light": 10, "three": [10, 38, 46], "tp": 10, "along": [10, 11, 13, 24, 38, 43], "manag": [10, 12, 16, 37, 41], "maintain": 10, "pipelin": [10, 30, 32, 37], "joint": [10, 19, 37, 39, 42], "independ": 10, "trial": 10, "been": [10, 14, 22, 39], "implement": [10, 13, 22, 24, 26, 27, 28, 36, 37, 38], "randomli": 10, "ampl": 10, "simpl": [11, 19, 38, 39, 42], "interfac": [11, 17, 20, 22, 26, 39, 41], "introduc": [11, 22, 35], "newoptimizationtrick": 11, "next": [11, 14, 19], "_default_config": 11, "take": [11, 21, 30, 36, 37, 38, 39, 40, 43, 44, 46], "passconfigparam": 11, "dataclass": 11, "hold": 11, "accelerator_typ": 11, "cudaexecutionprovid": [11, 22, 42, 47], "belong": 11, "dml": 11, "sinc": [11, 14, 19, 22, 36, 46, 47], "type_": 11, "categori": 11, "upload": [11, 12, 14, 19], "normal": [11, 30, 47], "root": [11, 12, 13, 45], "searchparamet": 11, "classmethod": 11, "cl": 11, "param1": 11, "param2": 11, "param3": 11, "objectclass": 11, "param4": 11, "union": [11, 37], "paramcategori": 11, "param5": 11, "param6": 11, "_run_for_config": 11, "engin": [11, 12, 19, 21, 25, 28, 35, 36, 40, 42, 47], "while": [11, 20, 22, 26, 36], "receiv": [11, 18, 37], "output_model_path": 11, "outlin": [12, 16], "discov": [12, 16], "asset": [12, 19, 30, 37, 41], "within": [12, 16], "aml": [12, 16, 19, 37, 41], "model_nam": [12, 13, 19, 43, 45], "model_vers": [12, 19], "my_subscription_id": [12, 45], "my_resource_group": [12, 45], "my_workspac": [12, 45], "my_datastor": [12, 45], "model_dir": [12, 45], "my_model": [12, 45], "pt": [12, 13, 38, 39, 45], "job_id": [12, 45], "my_job_id": [12, 45], "my_output_nam": [12, 45], "data_dir": [12, 37, 43, 44], "aml_system": [12, 16, 30, 37], "cluster": [12, 30, 35, 37, 47], "mcr": [12, 37, 47], "openmpi4": [12, 37, 47], "ubuntu20": [12, 37, 47], "04": [12, 37, 47], "conda": [12, 31, 37, 47], "yaml": [12, 13, 17, 37, 47], "Then": [12, 16, 30, 32, 34, 40, 43], "azuremlmodel": 12, "azuremldata": 12, "instruct": [12, 16, 28, 30, 36, 40], "setup": [12, 13, 32, 34, 38], "kubernet": [12, 35], "offer": [12, 36], "assist": [12, 37], "variou": [13, 15, 22, 36, 38], "usag": [13, 16, 24, 35], "package_config": 13, "run_config": 13, "tempdir": [13, 38], "advanc": [13, 35], "locat": [13, 14, 37, 41], "individu": [13, 39], "proprietari": [13, 38], "privat": [13, 38], "tempfil": 13, "peft": [13, 17, 22, 24], "model_name_or_path": 13, "trust_remote_cod": [13, 43], "use_ort_genai": 13, "d": [13, 17, 26, 41], "data_nam": [13, 16, 42, 43], "data_fil": [13, 43], "train_split": [13, 17], "eval_split": [13, 17], "text_field": 13, "text_templ": [13, 17], "max_seq_len": [13, 16], "output_path": [13, 17], "clean": [13, 37], "trust": [13, 43], "onnxrunti": 13, "split": [13, 16, 42, 43], "templat": [13, 37, 42], "question": [13, 17], "prompt": [13, 17, 43], "answer": [13, 17], "huggingface_model_optim": 13, "login": [13, 14], "extractedadapt": 13, "safetensor": 13, "pack_weight": [13, 17], "quantize_int4": [13, 17], "int4_quantization_mod": [13, 17], "Will": 13, "arrai": 13, "blockwis": 13, "py_vers": [13, 25, 28], "15": [13, 28], "delet": [13, 41], "compute_nam": [13, 41], "vm_size": [13, 41], "min_nod": [13, 41], "max_nod": [13, 41], "idle_time_before_scale_down": [13, 41], "vm": [13, 41], "idl": [13, 41], "scaledown": 13, "120": [13, 41], "simpli": [13, 32, 34, 43], "onnx_model_path": 13, "hf_model_nam": 13, "device_typ": 13, "data_config_path": 13, "predict_with_kv_cach": 13, "seq_len": [13, 43], "past_seq_len": [13, 43], "shared_kv": [13, 37], "ort_past_key_nam": [13, 37], "ort_past_value_nam": [13, 37], "max_sampl": [13, 42, 43], "fields_no_batch": 13, "disable_force_evaluate_other_ep": 13, "perf_tuning_output": 13, "perf_tun": [13, 16], "share": [13, 14, 37], "kv": 13, "sensit": 13, "key1": 13, "value1": 13, "key2": 13, "value2": 13, "account": [13, 14, 16, 37], "model_hash": 13, "hash": 13, "whenev": 14, "With": [14, 16, 27, 31, 36, 42], "easili": [14, 36, 37, 43], "nativ": [14, 18], "pip": [14, 23, 30, 38, 47], "ll": 14, "readi": [14, 18, 36, 39], "page": [14, 16, 40], "typic": [14, 18, 38, 47], "look": [14, 38, 39, 42, 43], "account_url": 14, "net": 14, "sure": [14, 16, 40], "az": 14, "enable_cloud_cach": 14, "container_nam": 14, "upload_to_cloud": 14, "olivecachemodel": 14, "immedi": 14, "skip": [14, 20, 37], "proce": 14, "until": 14, "complet": [14, 16, 24, 37, 38, 39, 43], "faster": [14, 25, 26, 27], "slower": 14, "influenc": 14, "internet": 14, "mani": [14, 20], "form": [15, 22, 25, 37], "torchtrtconvers": 15, "hfmodel": [16, 21, 24, 37, 42], "input_model": [16, 37, 38, 39, 42, 45], "meta": [16, 17, 37, 43], "7b": [16, 17, 37, 43], "catalog": 16, "futur": 16, "simplifi": [16, 17, 36, 47], "fetch": 16, "onnxconvers": [16, 20, 37, 38, 39, 42], "onnxconfig": 16, "manual": [16, 42], "input_id": [16, 37, 43], "attention_mask": [16, 43], "position_id": 16, "int64": [16, 22, 43], "dynamic_ax": [16, 37, 39], "sequence_length": [16, 37], "total_sequence_length": 16, "huggingfacecontain": [16, 42, 43], "oasst1_train": 16, "timdettm": 16, "openassist": 16, "guanaco": 16, "pre_process_data_config": [16, 42, 43], "text_col": 16, "512": 16, "pad_to_max_len": 16, "ortperftun": [16, 22, 37, 42], "huggingface_metr": [16, 37, 42], "f1": [16, 37, 42], "access": [16, 37, 40], "cli": 16, "termin": [16, 30], "vault": 16, "my_keyvault_nam": 16, "reserv": 16, "purpos": 16, "servic": 16, "msi": 16, "grant": 16, "guid": [16, 38, 39], "abov": [16, 17, 20, 24, 32, 34, 37, 42, 43, 47], "ptq": [16, 18, 33], "standard": [17, 39], "finetun": [17, 24], "usual": [17, 26], "power": [17, 26, 30, 42], "ux": 17, "v": 17, "flexibl": [17, 20, 22, 26, 30, 42], "describ": [17, 22, 24, 38, 44, 45, 46], "llama2_qlora": 17, "altern": [17, 19, 20, 44], "nampdn": 17, "tini": 17, "4096": 17, "4224": 17, "per_device_train_batch_s": 17, "per_device_eval_batch_s": 17, "max_step": 17, "150": 17, "logging_step": 17, "programming_languag": 17, "consumpt": 17, "known": [18, 19, 28, 38], "goe": 18, "through": [18, 22, 24, 26, 37, 42], "seri": 18, "product": 18, "design": [18, 35], "lora": [18, 22, 35], "loftq": 18, "mergeadapterweight": [18, 37], "There": [19, 26, 35, 37, 43, 47], "samplecod": 19, "onnxruntimepackag": 19, "models_rank": 19, "top": [19, 30, 37, 45], "across": [19, 22, 26, 36], "k": 19, "bestcandidatemodel_1": 19, "bestcandidatemodel_2": 19, "bestcandidatemodel_k": 19, "metric_1": 19, "metric_2": 19, "metric_3": 19, "sort": 19, "lower": [19, 24], "bestcandidatemodel": 19, "info": [19, 37], "model_config": 19, "onnxmodel": 19, "provider_opt": 19, "session_opt": [19, 37], "execution_mod": 19, "inter_op_num_thread": 19, "intra_op_num_thread": 19, "model_attributes_kei": 19, "model_attributes_valu": 19, "8602941176470589": 19, "higher_is_bett": [19, 37, 44], "36": 19, "2313": 19, "packaging_config_nam": 19, "model_rank": 19, "olive_output_model": 19, "olive_output_model_cpu": 19, "cpu_1": 19, "ml": [19, 25, 30, 35, 36, 45, 47], "No": 19, "endpoint": 19, "doesn": [19, 45, 47], "deploi": [19, 23, 40], "packagingtyp": 19, "outputmodel": 19, "export_in_mlflow_format": 19, "mlflow": 19, "registr": 19, "model_packag": 19, "modelpackageconfig": 19, "target_environ": 19, "target_environment_vers": 19, "plu": 19, "inferencing_serv": 19, "inferenceserverconfig": 19, "inferenc": [19, 22], "server": 19, "azuremlonlin": 19, "azuremlbatch": 19, "code_fold": 19, "score": 19, "scoring_script": 19, "base_environment_id": 19, "env": 19, "var": 19, "necessarili": 19, "deployment_config": 19, "deploymentconfig": 19, "endpoint_nam": 19, "deployment_nam": 19, "instance_typ": 19, "managedonlinedeploy": 19, "batchdeploy": 19, "interf": 19, "mini_batch_s": 19, "mini": 19, "invoc": 19, "10": [19, 37], "txt": [19, 25, 32, 34, 38, 43, 47], "include_runtime_packag": 19, "common_evalu": [19, 37, 42, 43], "local_system": [19, 37, 39, 42, 47], "null": [19, 37, 43], "53fc6781998a4624b61959bb064622c": 19, "0_onnxconvers": 19, "7a320d6d630bced3548f242238392730": 19, "1_orttransformersoptim": 19, "c499e39e42693aaab050820afd31e0c3": 19, "2_onnxquant": 19, "1431c563dcfda9c9c3bf26c5d61ef58": 19, "3_ortperftun": 19, "a843d77ae4964c04e145b83567fb5b05": 19, "cross": [20, 22, 23, 26], "onnxopversionconvers": [20, 37], "similar": 20, "modelbuild": [20, 37], "decreas": 20, "onnxfloattofloat16": 20, "often": 20, "expos": 20, "stage": 20, "stand": [20, 23], "alon": 20, "suit": 20, "architectur": [20, 25], "reli": 20, "grain": 20, "webgpu": 20, "onnxiofloat16tofloat32": 20, "give": [20, 35, 37], "poor": 20, "ortmixedprecis": 20, "minim": 20, "retain": 20, "skiplayernorm": 20, "fastgelu": 20, "embedlayernorm": 20, "mobil": 20, "dynamictofixedshap": [20, 37], "effici": [20, 23, 24, 26, 36], "coreml": 20, "helper": 20, "exclus": [20, 41], "deep": [21, 23, 24, 26, 27, 28, 43], "trtmodul": 21, "torch_tensorrt": 21, "regular": 21, "onnxmodeloptim": [22, 37], "involv": 22, "cost": [22, 26], "analyz": 22, "identifi": [22, 38], "argmax": 22, "isn": 22, "tensorproto": 22, "yet": 22, "orttransformersoptim": [22, 37, 42], "capabl": 22, "These": [22, 26, 39, 47], "done": 22, "appendprepostprocessingop": 22, "superresolut": 22, "output_format": 22, "png": 22, "use_audio_decod": 22, "leverag": [22, 43], "placehold": 22, "__model_input__": 22, "__model_output__": 22, "iomapentri": 22, "full": [22, 38, 46, 47], "ycbcrtopixel": 22, "bgr": 22, "io_map": 22, "y1_uint8": 22, "cb1_uint8": 22, "cr1_uint8": 22, "knowledg": [22, 42], "built": [22, 37, 42], "convertimagetobgr": 22, "incompat": 22, "web": 22, "kind": [22, 40, 47], "proper": [22, 24], "exactli": 22, "resiz": 22, "resize_to": 22, "input_index": 22, "dim_index": 22, "centercrop": 22, "height": 22, "width": [22, 24], "pixelstoycbcr": 22, "imagebytestofloat": 22, "unsqueez": 22, "ax": [22, 37], "squeez": 22, "floattoimagebyt": 22, "output_index": 22, "hw": 22, "multipli": 22, "convertbgrtoimag": 22, "image_format": 22, "insertbeamsearch": 22, "chain": [22, 36], "high": 22, "cover": [22, 43], "knob": 22, "perf_tuning_data_config": 22, "device_id": 22, "arena_extend_strategi": 22, "knextpoweroftwo": 22, "gpu_mem_limit": 22, "2147483648": 22, "cudnn_conv_algo_search": 22, "do_copy_in_default_stream": 22, "calib_data_config": [22, 23, 26, 39, 46], "small": [22, 37], "matric": [22, 24], "call": [22, 24, 26, 27, 39, 42, 43, 44], "extractadapt": [22, 37], "modifi": 22, "itself": 22, "add_initi": 22, "part": [22, 38], "As": [22, 26, 47], "visual": 23, "focus": [23, 26], "anywher": 23, "openvinoconvers": [23, 37], "ir": [23, 37, 42], "uniform": 23, "move": 23, "help": [23, 41, 42], "memori": [23, 24, 26, 27], "arithmet": 23, "undergo": 23, "actual": 23, "transit": 23, "happen": 23, "openvinoquantizationwithaccuraci": 23, "validation_func": 23, "max_drop": 23, "drop_typ": 23, "rank": [24, 44], "freez": 24, "inject": 24, "decomposit": 24, "hftrainingargu": 24, "backpropag": 24, "frozen": 24, "pretrain": 24, "simultan": 24, "fake": 24, "quantizationawaretrain": [24, 37], "ptldatamodul": 24, "ptlmodul": 24, "gpt": 24, "shot": 24, "unstructur": 24, "175b": 24, "176b": 24, "neglig": 24, "increas": [24, 26, 37], "sparsif": 24, "smaller": 24, "orthogon": 24, "least": [24, 43], "signific": 24, "row": 24, "speedup": 24, "wikitext2": 24, "direct": 25, "technologi": 25, "softwar": [25, 28], "chipset": 25, "variabl": [25, 28, 37, 40, 47], "qnn_sdk_root": 25, "qnnconvers": 25, "qnnmodellibgener": 25, "qnncontextbinarygener": 25, "qnn_backend": 25, "higher": [26, 37], "compar": [26, 27], "ones": 26, "consolid": [26, 27], "onnxquant": [26, 37, 39, 46], "hyperparamet": [26, 27], "onnxdynamicquant": [26, 37, 38], "onnxstaticquant": [26, 37], "qnnpreprocess": 26, "unsupport": 26, "olivepass": 26, "onnx_quant": [26, 37, 46], "popular": [26, 43], "mxnet": 26, "varieti": [26, 39], "distil": 26, "experi": [26, 36], "driven": 26, "friendli": [26, 27], "easi": [26, 27, 36, 42], "strive": 26, "reach": 26, "everywher": 26, "incquant": [26, 37], "incdynamicquant": [26, 37], "incstaticquant": [26, 37], "inc_quant": 26, "complex": 26, "less": 26, "bandwidth": 26, "vitisaiquant": [26, 37], "vitis_ai_quant": 26, "favorit": 27, "come": 27, "drop": 27, "gptqquantiz": [27, 37], "trade": 27, "wikitext2_train": 27, "3x": 27, "wa": 27, "upon": 27, "work": [27, 36, 37], "mit": 27, "autoawqquant": [27, 37], "snapdragon": 28, "hexagon": 28, "dsp": 28, "v2": 28, "18": 28, "240101": 28, "snpe_root": 28, "snpeconvers": [28, 37], "299": 28, "inceptionv3": 28, "predict": [28, 37, 43], "reshape_1": 28, "1001": 28, "snpequant": [28, 37], "represent": 28, "32bit": 28, "bias": 28, "quant_data_config": 28, "quarot": 29, "awai": 30, "enough": 30, "shut": 30, "interrupt": 30, "workflow_host": [30, 37], "configu": 30, "config_fil": [30, 42], "submit": 30, "safe": 30, "close": 30, "artifact": [30, 35, 37, 47], "yyyi": 30, "mm": 30, "dd": 30, "hh": 30, "ss": 30, "date": 30, "virtual": [31, 47], "public": 31, "pypi": 31, "git": 31, "egg": 31, "contribut": 31, "clone": [31, 38], "repositori": [31, 32, 34], "cd": 31, "tuner": [32, 34], "go": [32, 34], "quickstart": [32, 34, 35], "accord": [32, 34], "olive_run": [32, 34, 38, 39], "bert_auto": 32, "prepare_model_data": 34, "resnet_ptq_cpu": 34, "under": [35, 37, 42, 43], "releas": 35, "feel": 35, "free": [35, 40], "feedback": 35, "quick": [35, 39, 42], "tour": 35, "remot": [35, 38, 43, 47], "olivemodel": [35, 39], "searchalgorithm": 35, "industri": 36, "lead": [36, 39], "edg": 36, "consider": 36, "vendor": 36, "fragment": 36, "effort": 36, "toolchain": 36, "aim": 36, "aggreg": 36, "autom": 36, "unifi": 36, "serv": [36, 43], "plugin": 36, "innov": 36, "e2": 36, "solut": 36, "organ": [36, 37, 40, 43], "overview": [36, 40], "introduct": 36, "guidanc": 36, "tutori": [36, 38, 40], "vscode": 37, "intellisens": 37, "yml": 37, "suppurt": 37, "authent": 37, "credenti": 37, "too": 37, "3000": 37, "sometim": [37, 43], "hang": 37, "stream": 37, "issu": 37, "fail": 37, "exclude_managed_identity_credenti": 37, "4000": 37, "item": [37, 42, 43], "insensit": [37, 46], "Its": 37, "hub": [37, 45], "equival": [37, 38], "causal": 37, "try": [37, 42], "index": 37, "string_to_int_dim_param": 37, "kv_cach": [37, 43], "ort_present_key_nam": 37, "ort_present_value_nam": 37, "world_siz": 37, "num_hidden_lay": 37, "num_attention_head": 37, "past_sequence_length": 37, "sequence_length_idx": 37, "past_kv_dynamic_axi": 37, "present_kv_dynamic_axi": 37, "from_pretrain": 37, "tabl": [37, 43], "alia": [37, 47], "uniqu": [37, 43], "among": 37, "torchmetr": 37, "equal": 37, "metric_config": [37, 39, 44], "alter": 37, "subtyp": 37, "accuracysubtyp": [37, 44], "latencysubtyp": [37, 39, 44], "throughputsubtyp": [37, 44], "load_param": 37, "word_length": 37, "compute_param": 37, "pred": [37, 44], "result_kei": 37, "9": 37, "boolean": [37, 47], "percent": [37, 42, 44], "evaluate_func_kwarg": [37, 44], "signatur": 37, "posit": [37, 43], "metric_func_kwarg": 37, "pars": 37, "our": 37, "cifar": 37, "resourcegroup": 37, "create_dataload": [37, 44], "20": [37, 42, 44], "definit": 37, "pass_flow": [37, 42], "decid": [37, 46], "overridden": [37, 46], "onnxtransformersoptim": 37, "snpetoonnxconvers": 37, "xilinx": 37, "On": 37, "optimumconvers": 37, "optimummerg": 37, "mixedprecisionoverrid": 37, "onnx_convers": [37, 38], "calib_data_coonfig": 37, "transformers_optim": 37, "768": 37, "qualifi": 37, "stop": 37, "met": 37, "plot": 37, "verbos": [37, 38], "warn": 37, "error": [37, 47], "fatal": 37, "ort_log_severity_level": 37, "ort_py_log_severity_level": 37, "focu": 38, "simplest": 38, "my_model_acceleration_descript": 38, "local_requir": 38, "remote_requir": 38, "olive_config": 38, "launch": 38, "temporari": 38, "reason": 38, "disk": 38, "permiss": 38, "cuda_visible_devic": 38, "know": 38, "term": 38, "slightli": 38, "pytorchmodel": [38, 39, 45], "user_provided_info": 39, "user_provided_info_json_fil": 39, "loader": 39, "parse_obj": 39, "metrictyp": [39, 44], "latency_metr": [39, 44], "warmup_num": 39, "repeat_test_num": 39, "sleep_num": 39, "latency_data_config": [39, 44], "engine_config": 39, "onnx_conversion_config": 39, "quantization_config": 39, "best_execut": 39, "default_cpu_acceler": 39, "k8": 40, "home": [40, 47], "kube": 40, "kubeconfig": 40, "lunch": 40, "infrastructur": 40, "minut": 40, "verif": 40, "attach": 40, "coupl": 41, "indic": [41, 43, 47], "mutual": 41, "aml_config": 41, "redund": [42, 46], "At": 42, "still": [42, 45], "stack": 42, "disable_auto_optim": 42, "bert_gpu": 42, "auto_optimizer_config": 42, "tensorrtexecutionprovid": [42, 47], "trt_fp16": 42, "flow": 42, "catch": 42, "conveni": 42, "comparison": 42, "uncas": [42, 43], "mrpc": [42, 43], "classif": [42, 43], "glue": [42, 43], "subset": [42, 43], "input_col": [42, 43], "sentence1": [42, 43], "sentence2": [42, 43], "label_col": [42, 43], "label": [42, 43, 47], "cuda_transformers_optim": 42, "trt_transformers_optim": 42, "cuda_perf_tun": 42, "trt_perf_tun": 42, "enable_trt_fp16": 42, "rule": 42, "expert": 42, "open_llama_sparsegpt_gpu": 43, "dataset_1": 43, "dataset_2": 43, "letter": 43, "underscor": 43, "common_pass": 43, "dive": 43, "dummydatacontain": 43, "dummy_data_config_templ": 43, "token_type_id": 43, "datacomponentconfig": [43, 44], "huggingface_data_config": 43, "rawdatacontain": 43, "raw_data": 43, "input_dir": 43, "input_suffix": 43, "input_order_fil": 43, "input_ord": 43, "transformersdummydatacontain": 43, "transformers_dummy_data_config": 43, "config_path": 43, "transformerspromptdummydatacontain": 43, "transformerstokendummydatacontain": 43, "datacontain": 43, "four": 43, "load_dataset": 43, "local_dataset": 43, "simple_dataset": 43, "huggingface_dataset": 43, "raw_dataset": 43, "pre_process_data": 43, "pre_process": 43, "huggingface_pre_process": 43, "ner_huggingface_preprocess": 43, "text_generation_huggingface_pre_process": 43, "post_process_data": 43, "text_classification_post_process": 43, "ner_post_process": 43, "text_generation_post_process": 43, "default_dataload": 43, "no_auto_batch_dataload": 43, "show": 43, "rewrit": 43, "But": 43, "user_dir": 43, "customized_huggingface_dataset": 43, "register_dataset": 43, "_output": 43, "thing": [43, 46], "notic": 43, "register_pre_process": 43, "customized_huggingface_pre_process": 43, "customized_post_process": 43, "customized_dataload": 43, "inception_post_process": 43, "l8": 43, "l10": 43, "json_config": 43, "inception_config": 43, "l14": 43, "l16": 43, "conflict": 43, "metricgo": 44, "accuracy_metr": 44, "throughput_data_config": 44, "throughput_metr": 44, "accuracy_custom": 44, "eval_accuraci": 44, "yourself": 44, "model_output": 44, "omit": [44, 45], "my_model_dir": 45, "curat": 45, "fulli": 46, "olive_pass": 46, "create_pass_from_dict": 46, "mandatori": 47, "rais": 47, "localtargetuserconfig": 47, "prepend": 47, "python_system": 47, "virtualenv": 47, "myenv": 47, "openvinoexecutionprovid": 47, "python_environment_system": 47, "dmlexecutionprovid": 47, "pythonenvironmenttargetuserconfig": 47, "docker_system": 47, "mnist_requir": 47, "count": 47, "propag": 47, "cuda11": 47, "cudnn8": 47, "ubuntu22": 47, "filter": 47, "therefor": 47, "azureml_system": 47, "azuremltargetuserconfig": 47, "relev": 47, "isolatedorttargetuserconfig": 47}, "objects": {"olive.azureml.azureml_client": [[5, 0, 1, "", "AzureMLClientConfig"]], "olive.azureml.azureml_client.AzureMLClientConfig": [[5, 1, 1, "", "aml_config_path"], [5, 2, 1, "", "create_client"], [5, 2, 1, "", "create_registry_client"], [5, 1, 1, "", "default_auth_params"], [5, 2, 1, "", "get_workspace_config"], [5, 1, 1, "", "keyvault_name"], [5, 1, 1, "", "max_operation_retries"], [5, 1, 1, "", "operation_retry_interval"], [5, 1, 1, "", "read_timeout"], [5, 1, 1, "", "resource_group"], [5, 1, 1, "", "subscription_id"], [5, 1, 1, "", "workspace_name"]], "olive.engine": [[0, 3, 1, "", "Engine"], [0, 0, 1, "", "EngineConfig"]], "olive.engine.Engine": [[0, 2, 1, "", "register"], [0, 2, 1, "", "run"]], "olive.engine.EngineConfig": [[0, 1, 1, "", "cache_dir"], [0, 1, 1, "", "clean_cache"], [0, 1, 1, "", "clean_evaluation_cache"], [0, 1, 1, "", "evaluator"], [0, 1, 1, "", "host"], [0, 1, 1, "", "plot_pareto_frontier"], [0, 1, 1, "", "search_strategy"], [0, 1, 1, "", "target"]], "olive.evaluator.metric": [[2, 3, 1, "", "AccuracySubType"], [2, 3, 1, "", "LatencySubType"], [2, 0, 1, "", "Metric"], [2, 0, 1, "", "MetricGoal"], [2, 3, 1, "", "MetricType"], [2, 3, 1, "", "ThroughputSubType"]], "olive.evaluator.metric.AccuracySubType": [[2, 4, 1, "", "ACCURACY_SCORE"], [2, 4, 1, "", "AUROC"], [2, 4, 1, "", "F1_SCORE"], [2, 4, 1, "", "PERPLEXITY"], [2, 4, 1, "", "PRECISION"], [2, 4, 1, "", "RECALL"]], "olive.evaluator.metric.LatencySubType": [[2, 4, 1, "", "AVG"], [2, 4, 1, "", "MAX"], [2, 4, 1, "", "MIN"], [2, 4, 1, "", "P50"], [2, 4, 1, "", "P75"], [2, 4, 1, "", "P90"], [2, 4, 1, "", "P95"], [2, 4, 1, "", "P99"], [2, 4, 1, "", "P999"]], "olive.evaluator.metric.Metric": [[2, 1, 1, "", "backend"], [2, 1, 1, "", "data_config"], [2, 2, 1, "", "get_inference_settings"], [2, 2, 1, "", "get_run_kwargs"], [2, 2, 1, "", "get_sub_type_info"], [2, 1, 1, "", "name"], [2, 1, 1, "", "sub_types"], [2, 1, 1, "", "type"], [2, 1, 1, "", "user_config"]], "olive.evaluator.metric.MetricGoal": [[2, 2, 1, "", "has_regression_goal"], [2, 1, 1, "", "type"], [2, 1, 1, "", "value"]], "olive.evaluator.metric.MetricType": [[2, 4, 1, "", "ACCURACY"], [2, 4, 1, "", "CUSTOM"], [2, 4, 1, "", "LATENCY"], [2, 4, 1, "", "THROUGHPUT"]], "olive.evaluator.metric.ThroughputSubType": [[2, 4, 1, "", "AVG"], [2, 4, 1, "", "MAX"], [2, 4, 1, "", "MIN"], [2, 4, 1, "", "P50"], [2, 4, 1, "", "P75"], [2, 4, 1, "", "P90"], [2, 4, 1, "", "P95"], [2, 4, 1, "", "P99"], [2, 4, 1, "", "P999"]], "olive.evaluator.olive_evaluator": [[1, 3, 1, "", "OliveEvaluator"], [0, 0, 1, "", "OliveEvaluatorConfig"]], "olive.evaluator.olive_evaluator.OliveEvaluatorConfig": [[0, 2, 1, "", "create_evaluator"], [0, 5, 1, "", "is_accuracy_drop_tolerance"], [0, 1, 1, "", "metrics"], [0, 1, 1, "", "script_dir"], [0, 1, 1, "", "type"], [0, 1, 1, "", "type_args"], [0, 1, 1, "", "user_script"]], "olive.hardware.accelerator": [[7, 3, 1, "", "Device"]], "olive.hardware.accelerator.Device": [[7, 4, 1, "", "CPU"], [7, 4, 1, "", "CPU_SPR"], [7, 4, 1, "", "GPU"], [7, 4, 1, "", "INTEL_MYRIAD"], [7, 4, 1, "", "NPU"], [7, 4, 1, "", "VPU"]], "olive.model": [[3, 3, 1, "", "CompositeModelHandler"], [3, 3, 1, "", "DistributedHfModelHandler"], [3, 3, 1, "", "DistributedOnnxModelHandler"], [3, 3, 1, "", "HfModelHandler"], [3, 3, 1, "", "ModelConfig"], [3, 3, 1, "", "ONNXModelHandler"], [3, 3, 1, "", "OpenVINOModelHandler"], [3, 3, 1, "", "PyTorchModelHandler"], [3, 3, 1, "", "SNPEModelHandler"]], "olive.passes.pytorch.lora": [[4, 0, 1, "", "HFTrainingArguments"]], "olive.passes.pytorch.lora.HFTrainingArguments": [[4, 2, 1, "", "create_training_args"], [4, 1, 1, "", "evaluation_strategy"], [4, 1, 1, "", "extra_args"], [4, 1, 1, "", "gradient_checkpointing"], [4, 1, 1, "", "learning_rate"], [4, 1, 1, "", "lr_scheduler_type"], [4, 1, 1, "", "optim"], [4, 1, 1, "", "output_dir"], [4, 1, 1, "", "overwrite_output_dir"], [4, 1, 1, "", "report_to"], [4, 1, 1, "", "resume_from_checkpoint"], [4, 1, 1, "", "warmup_ratio"]], "olive.resource_path": [[5, 3, 1, "", "ResourceType"]], "olive.resource_path.ResourceType": [[5, 4, 1, "", "AzureMLDatastore"], [5, 4, 1, "", "AzureMLJobOutput"], [5, 4, 1, "", "AzureMLModel"], [5, 4, 1, "", "AzureMLRegistryModel"], [5, 4, 1, "", "LocalFile"], [5, 4, 1, "", "LocalFolder"], [5, 4, 1, "", "StringName"]], "olive.strategy.search_strategy": [[0, 3, 1, "", "SearchStrategy"], [0, 0, 1, "", "SearchStrategyConfig"]], "olive.strategy.search_strategy.SearchStrategyConfig": [[0, 1, 1, "", "execution_order"], [0, 1, 1, "", "max_iter"], [0, 1, 1, "", "max_time"], [0, 1, 1, "", "output_model_num"], [0, 1, 1, "", "search_algorithm"], [0, 1, 1, "", "search_algorithm_config"], [0, 1, 1, "", "stop_when_goals_met"]], "olive.systems.azureml": [[7, 0, 1, "", "AzureMLDockerConfig"], [7, 3, 1, "", "AzureMLSystem"]], "olive.systems.azureml.AzureMLDockerConfig": [[7, 1, 1, "", "base_image"], [7, 1, 1, "", "build_context_path"], [7, 1, 1, "", "conda_file_path"], [7, 1, 1, "", "dockerfile"], [7, 1, 1, "", "name"], [7, 1, 1, "", "version"]], "olive.systems.common": [[7, 3, 1, "", "AcceleratorConfig"], [7, 3, 1, "", "SystemType"]], "olive.systems.common.AcceleratorConfig": [[7, 1, 1, "", "device"], [7, 1, 1, "", "execution_providers"]], "olive.systems.common.SystemType": [[7, 4, 1, "", "AzureML"], [7, 4, 1, "", "Docker"], [7, 4, 1, "", "IsolatedORT"], [7, 4, 1, "", "Local"], [7, 4, 1, "", "PythonEnvironment"]], "olive.systems.docker": [[7, 3, 1, "", "DockerSystem"], [7, 0, 1, "", "LocalDockerConfig"]], "olive.systems.docker.LocalDockerConfig": [[7, 1, 1, "", "build_args"], [7, 1, 1, "", "build_context_path"], [7, 1, 1, "", "dockerfile"], [7, 1, 1, "", "image_name"], [7, 1, 1, "", "run_params"]], "olive.systems.isolated_ort": [[7, 3, 1, "", "IsolatedORTSystem"]], "olive.systems.local": [[7, 3, 1, "", "LocalSystem"]], "olive.systems.python_environment": [[7, 3, 1, "", "PythonEnvironmentSystem"]], "olive.systems": [[7, 6, 0, "-", "system_alias"]], "olive.systems.system_alias": [[7, 3, 1, "", "AzureND12SSystem"], [7, 3, 1, "", "AzureND24RSSystem"], [7, 3, 1, "", "AzureND24SSystem"], [7, 3, 1, "", "AzureND6SSystem"], [7, 3, 1, "", "AzureND96A100System"], [7, 3, 1, "", "AzureND96ASystem"], [7, 3, 1, "", "AzureNDV2System"], [7, 3, 1, "", "SurfaceProSystem1796"], [7, 3, 1, "", "SurfaceProSystem1807"]], "olive.systems.system_alias.AzureND12SSystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND24RSSystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND24SSystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND6SSystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND96A100System": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureND96ASystem": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.AzureNDV2System": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "num_cpus"], [7, 4, 1, "", "num_gpus"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.SurfaceProSystem1796": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_alias.SurfaceProSystem1807": [[7, 4, 1, "", "accelerators"], [7, 4, 1, "", "sku"], [7, 4, 1, "", "system_type"]], "olive.systems.system_config": [[7, 0, 1, "", "AzureMLTargetUserConfig"], [7, 0, 1, "", "DockerTargetUserConfig"], [7, 0, 1, "", "IsolatedORTTargetUserConfig"], [7, 0, 1, "", "LocalTargetUserConfig"], [7, 0, 1, "", "PythonEnvironmentTargetUserConfig"], [7, 0, 1, "", "SystemConfig"]], "olive.systems.system_config.AzureMLTargetUserConfig": [[7, 1, 1, "", "aml_compute"], [7, 1, 1, "", "aml_docker_config"], [7, 1, 1, "", "aml_environment_config"], [7, 1, 1, "", "azureml_client_config"], [7, 1, 1, "", "datastores"], [7, 1, 1, "", "instance_count"], [7, 1, 1, "", "is_dev"], [7, 1, 1, "", "olive_managed_env"], [7, 1, 1, "", "requirements_file"], [7, 1, 1, "", "resources"], [7, 1, 1, "", "tags"]], "olive.systems.system_config.DockerTargetUserConfig": [[7, 1, 1, "", "is_dev"], [7, 1, 1, "", "local_docker_config"], [7, 1, 1, "", "olive_managed_env"], [7, 1, 1, "", "requirements_file"]], "olive.systems.system_config.IsolatedORTTargetUserConfig": [[7, 1, 1, "", "environment_variables"], [7, 1, 1, "", "prepend_to_path"], [7, 1, 1, "", "python_environment_path"]], "olive.systems.system_config.LocalTargetUserConfig": [[7, 1, 1, "", "accelerators"], [7, 1, 1, "", "hf_token"]], "olive.systems.system_config.PythonEnvironmentTargetUserConfig": [[7, 1, 1, "", "olive_managed_env"], [7, 1, 1, "", "requirements_file"]], "olive.systems.system_config.SystemConfig": [[7, 1, 1, "", "config"], [7, 2, 1, "", "create_system"], [7, 5, 1, "", "olive_managed_env"], [7, 1, 1, "", "type"]], "": [[4, 7, 1, "cmdoption-arg-ActivationSymmetric", "ActivationSymmetric"], [4, 7, 1, "cmdoption-arg-AddQDQPairToWeight", "AddQDQPairToWeight"], [4, 7, 1, "cmdoption-arg-EnableSubgraph", "EnableSubgraph"], [4, 7, 1, "cmdoption-arg-ForceQuantizeNoInputCheck", "ForceQuantizeNoInputCheck"], [4, 7, 1, "cmdoption-arg-MatMulConstBOnly", "MatMulConstBOnly"], [4, 7, 1, "cmdoption-arg-WeightSymmetric", "WeightSymmetric"], [4, 7, 1, "cmdoption-arg-a_asym", "a_asym"], [4, 7, 1, "cmdoption-arg-a_bits", "a_bits"], [4, 7, 1, "cmdoption-arg-a_clip_ratio", "a_clip_ratio"], [4, 7, 1, "cmdoption-arg-a_groupsize", "a_groupsize"], [4, 7, 1, "cmdoption-arg-activation_type", "activation_type"], [4, 7, 1, "cmdoption-arg-all_tensors_to_one_file", "all_tensors_to_one_file"], [4, 7, 1, "cmdoption-arg-allow_tf32", "allow_tf32"], [4, 7, 1, "cmdoption-arg-append_first_op_types_to_quantize_list", "append_first_op_types_to_quantize_list"], [4, 7, 1, "cmdoption-arg-approach", "approach"], [4, 7, 1, "cmdoption-arg-atol", "atol"], [5, 7, 1, "cmdoption-arg-azureml_client", "azureml_client"], [4, 7, 1, "cmdoption-arg-backend", "backend"], [4, 7, 1, "cmdoption-arg-binary_file", "binary_file"], [4, 7, 1, "cmdoption-arg-bits", "bits"], [4, 7, 1, "cmdoption-arg-blocksize", "blocksize"], [4, 7, 1, "cmdoption-arg-calibrate_method", "calibrate_method"], [4, 7, 1, "cmdoption-arg-calibration_batch_size", "calibration_batch_size"], [4, 7, 1, "cmdoption-arg-calibration_data_config", "calibration_data_config"], [4, 7, 1, "cmdoption-arg-calibration_nsamples", "calibration_nsamples"], [4, 7, 1, "cmdoption-arg-calibration_sampling_size", "calibration_sampling_size"], [4, 7, 1, "cmdoption-arg-checkpoint_path", "checkpoint_path"], [4, 7, 1, "cmdoption-arg-components", "components"], [4, 7, 1, "cmdoption-arg-compress_to_fp16", "compress_to_fp16"], [4, 7, 1, "cmdoption-arg-compute_dtype", "compute_dtype"], [4, 7, 1, "cmdoption-arg-convert_attribute", "convert_attribute"], [4, 7, 1, "cmdoption-arg-cpu_cores", "cpu_cores"], [4, 7, 1, "cmdoption-arg-damp_percent", "damp_percent"], [4, 7, 1, "cmdoption-arg-data_config", "data_config"], [5, 7, 1, "cmdoption-arg-datastore_name", "datastore_name"], [5, 7, 1, "cmdoption-arg-datastore_url", "datastore_url"], [4, 7, 1, "cmdoption-arg-desc_act", "desc_act"], [4, 7, 1, "cmdoption-arg-device", "device"], [4, 7, 1, "cmdoption-arg-dim_param", "dim_param"], [4, 7, 1, "cmdoption-arg-dim_value", "dim_value"], [4, 7, 1, "cmdoption-arg-do_validate", "do_validate"], [4, 7, 1, "cmdoption-arg-domain", "domain"], [4, 7, 1, "cmdoption-arg-double_quant", "double_quant"], [4, 7, 1, "cmdoption-arg-duo_scaling", "duo_scaling"], [4, 7, 1, "cmdoption-arg-dynamic_lora_r", "dynamic_lora_r"], [4, 7, 1, "cmdoption-arg-element_wise_binary_ops", "element_wise_binary_ops"], [4, 7, 1, "cmdoption-arg-enable_cuda_graph", "enable_cuda_graph"], [4, 7, 1, "cmdoption-arg-enable_dpu", "enable_dpu"], [4, 7, 1, "cmdoption-arg-enable_htp", "enable_htp"], [4, 7, 1, "cmdoption-arg-enable_profiling", "enable_profiling"], [4, 7, 1, "cmdoption-arg-eval_data_config", "eval_data_config"], [4, 7, 1, "cmdoption-arg-example_input_func", "example_input_func"], [4, 7, 1, "cmdoption-arg-exclude_embeds", "exclude_embeds"], [4, 7, 1, "cmdoption-arg-exclude_lm_head", "exclude_lm_head"], [4, 7, 1, "cmdoption-arg-excluded_precisions", "excluded_precisions"], [4, 7, 1, "cmdoption-arg-execution_mode_list", "execution_mode_list"], [4, 7, 1, "cmdoption-arg-export_compatible", "export_compatible"], [4, 7, 1, "cmdoption-arg-external_data_name", "external_data_name"], [4, 7, 1, "cmdoption-arg-extra_args", "extra_args"], [4, 7, 1, "cmdoption-arg-extra_configs", "extra_configs"], [4, 7, 1, "cmdoption-arg-extra_options", "extra_options"], [4, 7, 1, "cmdoption-arg-extra_session_config", "extra_session_config"], [4, 7, 1, "cmdoption-arg-final_orientation", "final_orientation"], [4, 7, 1, "cmdoption-arg-float16", "float16"], [4, 7, 1, "cmdoption-arg-force_evaluate_other_eps", "force_evaluate_other_eps"], [4, 7, 1, "cmdoption-arg-force_fp16_inputs", "force_fp16_inputs"], [4, 7, 1, "cmdoption-arg-force_fp32_nodes", "force_fp32_nodes"], [4, 7, 1, "cmdoption-arg-force_fp32_ops", "force_fp32_ops"], [4, 7, 1, "cmdoption-arg-fp16", "fp16"], [4, 7, 1, "cmdoption-arg-fuse_layernorm", "fuse_layernorm"], [4, 7, 1, "cmdoption-arg-gptq_damping", "gptq_damping"], [4, 7, 1, "cmdoption-arg-gptq_opt_scales", "gptq_opt_scales"], [4, 7, 1, "cmdoption-arg-gpus", "gpus"], [6, 7, 1, "cmdoption-arg-group", "group"], [4, 7, 1, "cmdoption-arg-group_size", "group_size"], [4, 7, 1, "cmdoption-arg-hidden_size", "hidden_size"], [4, 7, 1, "cmdoption-arg-htp_socs", "htp_socs"], [4, 7, 1, "cmdoption-arg-ignored_scope", "ignored_scope"], [4, 7, 1, "cmdoption-arg-ignored_scope_type", "ignored_scope_type"], [4, 7, 1, "cmdoption-arg-input", "input"], [4, 7, 1, "cmdoption-arg-input_dim", "input_dim"], [4, 7, 1, "cmdoption-arg-input_int32", "input_int32"], [4, 7, 1, "cmdoption-arg-input_layouts", "input_layouts"], [4, 7, 1, "cmdoption-arg-input_model_dtype", "input_model_dtype"], [4, 7, 1, "cmdoption-arg-input_name", "input_name"], [4, 7, 1, "cmdoption-arg-input_names", "input_names"], [4, 7, 1, "cmdoption-arg-input_nodes", "input_nodes"], [4, 7, 1, "cmdoption-arg-input_shape", "input_shape"], [4, 7, 1, "cmdoption-arg-input_shapes", "input_shapes"], [4, 7, 1, "cmdoption-arg-input_types", "input_types"], [4, 7, 1, "cmdoption-arg-inputs_to_make_channel_last", "inputs_to_make_channel_last"], [4, 7, 1, "cmdoption-arg-inside_layer_modules", "inside_layer_modules"], [4, 7, 1, "cmdoption-arg-int4_accuracy_level", "int4_accuracy_level"], [4, 7, 1, "cmdoption-arg-int4_block_size", "int4_block_size"], [4, 7, 1, "cmdoption-arg-inter_thread_num_list", "inter_thread_num_list"], [4, 7, 1, "cmdoption-arg-intra_thread_num_list", "intra_thread_num_list"], [4, 7, 1, "cmdoption-arg-io_bind", "io_bind"], [5, 7, 1, "cmdoption-arg-job_name", "job_name"], [4, 7, 1, "cmdoption-arg-k_bits", "k_bits"], [4, 7, 1, "cmdoption-arg-k_clip_ratio", "k_clip_ratio"], [4, 7, 1, "cmdoption-arg-k_groupsize", "k_groupsize"], [4, 7, 1, "cmdoption-arg-keep_io_types", "keep_io_types"], [4, 7, 1, "cmdoption-arg-layer_name_filter", "layer_name_filter"], [4, 7, 1, "cmdoption-arg-layers_block_name", "layers_block_name"], [4, 7, 1, "cmdoption-arg-lib_name", "lib_name"], [4, 7, 1, "cmdoption-arg-lib_targets", "lib_targets"], [4, 7, 1, "cmdoption-arg-loftq_iter", "loftq_iter"], [4, 7, 1, "cmdoption-arg-logger", "logger"], [4, 7, 1, "cmdoption-arg-lora_alpha", "lora_alpha"], [4, 7, 1, "cmdoption-arg-lora_dropout", "lora_dropout"], [4, 7, 1, "cmdoption-arg-lora_r", "lora_r"], [4, 7, 1, "cmdoption-arg-make_inputs", "make_inputs"], [4, 7, 1, "cmdoption-arg-max_finite_val", "max_finite_val"], [4, 7, 1, "cmdoption-arg-max_layer", "max_layer"], [4, 7, 1, "cmdoption-arg-merge_adapter_weights", "merge_adapter_weights"], [4, 7, 1, "cmdoption-arg-metadata_only", "metadata_only"], [4, 7, 1, "cmdoption-arg-metric", "metric"], [4, 7, 1, "cmdoption-arg-min_layer", "min_layer"], [4, 7, 1, "cmdoption-arg-min_positive_val", "min_positive_val"], [4, 7, 1, "cmdoption-arg-model_type", "model_type"], [4, 7, 1, "cmdoption-arg-modules_to_fuse", "modules_to_fuse"], [4, 7, 1, "cmdoption-arg-modules_to_not_convert", "modules_to_not_convert"], [4, 7, 1, "cmdoption-arg-modules_to_save", "modules_to_save"], [6, 7, 1, "cmdoption-arg-multivariate", "multivariate"], [5, 7, 1, "cmdoption-arg-name", "name"], [4, 7, 1, "cmdoption-arg-name_pattern", "name_pattern"], [4, 7, 1, "cmdoption-arg-need_layer_fusing", "need_layer_fusing"], [4, 7, 1, "cmdoption-arg-no_repeat_ngram_size", "no_repeat_ngram_size"], [4, 7, 1, "cmdoption-arg-node_block_list", "node_block_list"], [4, 7, 1, "cmdoption-arg-nodes_to_exclude", "nodes_to_exclude"], [4, 7, 1, "cmdoption-arg-nodes_to_quantize", "nodes_to_quantize"], [4, 7, 1, "cmdoption-arg-nsamples", "nsamples"], [4, 7, 1, "cmdoption-arg-num_epochs", "num_epochs"], [4, 7, 1, "cmdoption-arg-num_heads", "num_heads"], [4, 7, 1, "cmdoption-arg-num_key_value_heads", "num_key_value_heads"], [6, 7, 1, "cmdoption-arg-num_samples", "num_samples"], [4, 7, 1, "cmdoption-arg-num_steps", "num_steps"], [4, 7, 1, "cmdoption-arg-only_onnxruntime", "only_onnxruntime"], [4, 7, 1, "cmdoption-arg-op_block_list", "op_block_list"], [4, 7, 1, "cmdoption-arg-op_type_dict", "op_type_dict"], [4, 7, 1, "cmdoption-arg-op_types_to_quantize", "op_types_to_quantize"], [4, 7, 1, "cmdoption-arg-opt_level", "opt_level"], [4, 7, 1, "cmdoption-arg-opt_level_list", "opt_level_list"], [4, 7, 1, "cmdoption-arg-optimization_options", "optimization_options"], [4, 7, 1, "cmdoption-arg-optimize_model", "optimize_model"], [4, 7, 1, "cmdoption-arg-optional_inputs", "optional_inputs"], [4, 7, 1, "cmdoption-arg-out_node", "out_node"], [4, 7, 1, "cmdoption-arg-output_model", "output_model"], [5, 7, 1, "cmdoption-arg-output_name", "output_name"], [4, 7, 1, "cmdoption-arg-output_names", "output_names"], [4, 7, 1, "cmdoption-arg-output_nodes", "output_nodes"], [4, 7, 1, "cmdoption-arg-outputs_to_make_channel_last", "outputs_to_make_channel_last"], [4, 7, 1, "cmdoption-arg-outside_layer_modules", "outside_layer_modules"], [4, 7, 1, "cmdoption-arg-overrides_config", "overrides_config"], [4, 7, 1, "cmdoption-arg-pack_inputs", "pack_inputs"], [4, 7, 1, "cmdoption-arg-pack_model_for_onnx_conversion", "pack_model_for_onnx_conversion"], [4, 7, 1, "cmdoption-arg-parallel_jobs", "parallel_jobs"], [4, 7, 1, "cmdoption-arg-past_key_value_name", "past_key_value_name"], [5, 7, 1, "cmdoption-arg-path", "path"], [4, 7, 1, "cmdoption-arg-per_channel", "per_channel"], [4, 7, 1, "cmdoption-arg-percdamp", "percdamp"], [4, 7, 1, "cmdoption-arg-post", "post"], [4, 7, 1, "cmdoption-arg-pre", "pre"], [4, 7, 1, "cmdoption-arg-precision", "precision"], [4, 7, 1, "cmdoption-arg-prepare_qnn_config", "prepare_qnn_config"], [4, 7, 1, "cmdoption-arg-preset", "preset"], [4, 7, 1, "cmdoption-arg-providers_list", "providers_list"], [4, 7, 1, "cmdoption-arg-ptl_data_module", "ptl_data_module"], [4, 7, 1, "cmdoption-arg-ptl_module", "ptl_module"], [4, 7, 1, "cmdoption-arg-q_group_size", "q_group_size"], [4, 7, 1, "cmdoption-arg-qconfig_func", "qconfig_func"], [4, 7, 1, "cmdoption-arg-qnn_extra_options", "qnn_extra_options"], [4, 7, 1, "cmdoption-arg-quant_format", "quant_format"], [4, 7, 1, "cmdoption-arg-quant_level", "quant_level"], [4, 7, 1, "cmdoption-arg-quant_mode", "quant_mode"], [4, 7, 1, "cmdoption-arg-quant_preprocess", "quant_preprocess"], [4, 7, 1, "cmdoption-arg-quant_type", "quant_type"], [4, 7, 1, "cmdoption-arg-recipes", "recipes"], [4, 7, 1, "cmdoption-arg-reduce_range", "reduce_range"], [5, 7, 1, "cmdoption-arg-relative_path", "relative_path"], [4, 7, 1, "cmdoption-arg-rotate", "rotate"], [4, 7, 1, "cmdoption-arg-rotation_seed", "rotation_seed"], [4, 7, 1, "cmdoption-arg-round_interval", "round_interval"], [4, 7, 1, "cmdoption-arg-save_as_external_data", "save_as_external_data"], [4, 7, 1, "cmdoption-arg-save_format", "save_format"], [4, 7, 1, "cmdoption-arg-save_metadata_for_token_generation", "save_metadata_for_token_generation"], [4, 7, 1, "cmdoption-arg-script_dir", "script_dir"], [4, 7, 1, "cmdoption-arg-search", "search"], [4, 7, 1, "cmdoption-arg-seed", "seed"], [4, 7, 1, "cmdoption-arg-size_threshold", "size_threshold"], [4, 7, 1, "cmdoption-arg-sparsity", "sparsity"], [4, 7, 1, "cmdoption-arg-static_groups", "static_groups"], [4, 7, 1, "cmdoption-arg-strict", "strict"], [4, 7, 1, "cmdoption-arg-sym", "sym"], [4, 7, 1, "cmdoption-arg-target_device", "target_device"], [4, 7, 1, "cmdoption-arg-target_modules", "target_modules"], [4, 7, 1, "cmdoption-arg-target_opset", "target_opset"], [4, 7, 1, "cmdoption-arg-tool_command", "tool_command"], [4, 7, 1, "cmdoption-arg-tool_command_args", "tool_command_args"], [4, 7, 1, "cmdoption-arg-torch_dtype", "torch_dtype"], [4, 7, 1, "cmdoption-arg-train_data_config", "train_data_config"], [4, 7, 1, "cmdoption-arg-training_args", "training_args"], [4, 7, 1, "cmdoption-arg-training_loop_func", "training_loop_func"], [4, 7, 1, "cmdoption-arg-trt_fp16_enable", "trt_fp16_enable"], [4, 7, 1, "cmdoption-arg-true_sequential", "true_sequential"], [4, 7, 1, "cmdoption-arg-tuning_criterion", "tuning_criterion"], [4, 7, 1, "cmdoption-arg-use_dynamo_exporter", "use_dynamo_exporter"], [4, 7, 1, "cmdoption-arg-use_enhanced_quantizer", "use_enhanced_quantizer"], [4, 7, 1, "cmdoption-arg-use_external_data_format", "use_external_data_format"], [4, 7, 1, "cmdoption-arg-use_forced_decoder_ids", "use_forced_decoder_ids"], [4, 7, 1, "cmdoption-arg-use_gpu", "use_gpu"], [4, 7, 1, "cmdoption-arg-use_gqa", "use_gqa"], [4, 7, 1, "cmdoption-arg-use_logits_processor", "use_logits_processor"], [4, 7, 1, "cmdoption-arg-use_prefix_vocab_mask", "use_prefix_vocab_mask"], [4, 7, 1, "cmdoption-arg-use_symbolic_shape_infer", "use_symbolic_shape_infer"], [4, 7, 1, "cmdoption-arg-use_temperature", "use_temperature"], [4, 7, 1, "cmdoption-arg-use_vocab_mask", "use_vocab_mask"], [4, 7, 1, "cmdoption-arg-user_script", "user_script"], [4, 7, 1, "cmdoption-arg-v_bits", "v_bits"], [4, 7, 1, "cmdoption-arg-v_clip_ratio", "v_clip_ratio"], [4, 7, 1, "cmdoption-arg-v_groupsize", "v_groupsize"], [4, 7, 1, "cmdoption-arg-val_data_config", "val_data_config"], [4, 7, 1, "cmdoption-arg-version", "version"], [4, 7, 1, "cmdoption-arg-w_asym", "w_asym"], [4, 7, 1, "cmdoption-arg-w_bit", "w_bit"], [4, 7, 1, "cmdoption-arg-w_bits", "w_bits"], [4, 7, 1, "cmdoption-arg-w_gptq", "w_gptq"], [4, 7, 1, "cmdoption-arg-w_groupsize", "w_groupsize"], [4, 7, 1, "cmdoption-arg-w_rtn", "w_rtn"], [4, 7, 1, "cmdoption-arg-weight_only_config", "weight_only_config"], [4, 7, 1, "cmdoption-arg-weight_type", "weight_type"], [6, 7, 1, "cmdoption-arg-with_replacement", "with_replacement"], [4, 7, 1, "cmdoption-arg-workspace", "workspace"], [4, 7, 1, "cmdoption-arg-zero_point", "zero_point"]], "extra.Sigmoid": [[4, 7, 1, "cmdoption-arg-extra.Sigmoid.nnapi", "nnapi"]]}, "objtypes": {"0": "py:pydantic_settings", "1": "py:pydantic_field", "2": "py:method", "3": "py:class", "4": "py:attribute", "5": "py:property", "6": "py:module", "7": "std:cmdoption"}, "objnames": {"0": ["py", "pydantic_settings", "Python settings"], "1": ["py", "pydantic_field", "Python field"], "2": ["py", "method", "Python method"], "3": ["py", "class", "Python class"], "4": ["py", "attribute", "Python attribute"], "5": ["py", "property", "Python property"], "6": ["py", "module", "Python module"], "7": ["std", "cmdoption", "program option"]}, "titleterms": {"engin": [0, 10, 37, 38, 39], "engineconfig": 0, "searchstrategi": 0, "oliveevalu": 1, "metric": [2, 16, 19, 44], "metrictyp": 2, "accuracysubtyp": 2, "latencysubtyp": 2, "throughputsubtyp": 2, "metricgo": 2, "olivemodel": 3, "model": [3, 12, 13, 14, 15, 16, 18, 19, 20, 22, 23, 25, 28, 29, 32, 34, 37, 38, 39, 45], "configur": [3, 11, 13, 14, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 30, 42, 43, 44, 46, 47], "hf": 3, "handler": 3, "distribut": 3, "pytorch": [3, 21, 24, 27], "onnx": [3, 20, 22, 26], "openvino": [3, 23], "snpe": [3, 28], "composit": 3, "pass": [4, 10, 11, 13, 37, 38, 39, 46], "onnxconvers": 4, "onnxopversionconvers": 4, "onnxmodeloptim": 4, "orttransformersoptim": 4, "ortperftun": 4, "onnxfloattofloat16": 4, "onnxiofloat16tofloat32": 4, "ortmixedprecis": 4, "qnnpreprocess": 4, "mixedprecisionoverrid": 4, "onnxdynamicquant": 4, "onnxstaticquant": 4, "onnxquant": 4, "dynamictofixedshap": 4, "incdynamicquant": 4, "incstaticquant": 4, "incquant": 4, "vitisaiquant": 4, "appendprepostprocessingop": 4, "insertbeamsearch": 4, "extractadapt": 4, "lora": [4, 13, 17, 24], "qlora": [4, 24], "loftq": [4, 24], "hftrainingargu": 4, "quantizationawaretrain": 4, "openvinoconvers": 4, "openvinoquant": 4, "snpeconvers": 4, "snpequant": 4, "snpetoonnxconvers": 4, "qnnconvers": 4, "qnnmodellibgener": 4, "qnncontextbinarygener": 4, "mergeadapterweight": [4, 24], "sparsegpt": [4, 24], "slicegpt": [4, 24], "quarot": [4, 27], "gptqquantiz": 4, "autoawqquant": 4, "torchtrtconvers": [4, 21], "optimumconvers": 4, "optimummerg": 4, "modelbuild": 4, "resourcepath": 5, "resourcetyp": 5, "localfil": 5, "localfold": 5, "stringnam": 5, "azuremlmodel": [5, 19], "azuremldatastor": 5, "azuremljoboutput": 5, "searchalgorithm": 6, "exhaustivesearchalgorithm": 6, "randomsearchalgorithm": 6, "tpesearchalgorithm": 6, "olivesystem": 7, "config": [7, 13, 16, 19, 32, 34, 37, 43], "systemtyp": 7, "devic": [7, 12], "acceleratorconfig": [7, 47], "localtargetuserconfig": 7, "dockertargetuserconfig": 7, "azuremltargetuserconfig": 7, "pythonenvironmenttargetuserconfig": 7, "isolatedorttargetuserconfig": 7, "class": [7, 11], "localsystem": 7, "azuremlsystem": 7, "dockersystem": 7, "pythonenvironmentsystem": 7, "isolatedortsystem": 7, "system": [7, 10, 16, 30, 37, 39, 47], "alia": 7, "exampl": [8, 9, 11, 16, 20, 21, 22, 23, 24, 25, 26, 27, 28, 33, 35, 37], "custom": [9, 43, 44], "script": [9, 12, 41], "user_script": 9, "script_dir": 9, "design": 10, "evalu": [10, 37, 39], "search": [10, 22], "strategi": 10, "execut": 10, "order": 10, "algorithm": 10, "how": [11, 14, 19, 42, 43, 44, 45, 46, 47], "add": [11, 40], "new": 11, "optim": [11, 18, 22, 32, 34, 42], "1": [11, 23], "defin": 11, "2": [11, 23], "3": 11, "implement": 11, "run": [11, 13, 30, 32, 34, 39], "function": 11, "azur": [12, 14, 16, 30, 37, 40, 41], "ml": [12, 16, 37, 41], "integr": [12, 16], "machin": [12, 30, 40], "learn": [12, 30, 40], "client": [12, 37], "oliv": [12, 13, 19, 23, 25, 28, 30, 35, 36, 37, 38], "core": 12, "us": [12, 14, 32, 34], "azureml": [12, 16, 30, 37, 45, 47], "regist": [12, 39], "curat": 12, "store": 12, "datastor": [12, 45], "from": [12, 23, 31], "an": [12, 30], "job": [12, 45], "output": [12, 20, 30, 45], "data": [12, 13, 34, 43], "comput": [12, 13, 30], "host": [12, 37, 39, 40], "target": [12, 39], "packag": [12, 19], "workspac": [12, 30], "connect": 12, "your": 12, "own": 12, "arc": [12, 40], "helper": 12, "command": 13, "line": 13, "tool": [13, 23], "sub": 13, "name": [13, 45], "argument": 13, "finetun": 13, "option": [13, 23, 31, 37], "dataset": [13, 16], "remot": [13, 30, 45], "export": 13, "adapt": [13, 17, 22], "int4": 13, "quantiz": [13, 23, 24, 25, 26, 28, 29], "qualcomm": 13, "sdk": [13, 25, 28], "manag": [13, 47], "aml": [13, 40], "tune": [13, 22], "session": 13, "param": 13, "which": 13, "mutual": 13, "exclus": 13, "huggingfac": [13, 16], "ar": [13, 32], "provid": 13, "user": [13, 39], "should": 13, "follow": 13, "modifi": 13, "default": 13, "pleas": 13, "refer": [13, 35], "contain": 13, "transformerstokendummydatacontain": 13, "more": 13, "detail": 13, "cloud": [13, 14], "cach": [13, 14], "what": [14, 19], "i": [14, 19], "instal": [14, 23, 30, 31], "extra": [14, 23, 30], "depend": [14, 30, 31], "blob": 14, "storag": 14, "note": 14, "convers": [15, 20, 23, 25, 28], "introduct": 16, "input": [16, 20, 37, 38, 39], "hub": 16, "local": [16, 40, 45, 47], "load": 16, "login": 16, "docker": [16, 47], "python": [16, 47], "environ": [16, 47], "e2": 16, "transform": [18, 22], "artifact": 19, "zipfil": 19, "candidatemodel": 19, "rank": 19, "json": [19, 32, 37], "file": [19, 45], "azuremldata": 19, "azuremldeploy": 19, "dockerfil": 19, "infer": 19, "float16": 20, "float32": 20, "mix": 20, "precis": 20, "convert": 20, "dynam": 20, "shape": 20, "fix": 20, "ort": [22, 47], "append": 22, "pre": 22, "post": [22, 23, 28], "process": 22, "op": 22, "insert": 22, "beam": 22, "perform": 22, "extract": 22, "prerequisit": [23, 25, 28, 32, 34, 47], "runtim": 23, "develop": 23, "pypi": 23, "train": [23, 24, 28], "ptq": [23, 28, 32, 34], "awar": 24, "qnn": 25, "download": [25, 28], "unzip": [25, 28], "librari": 25, "gener": [25, 43], "context": 25, "binari": 25, "onnxruntim": 26, "intel": 26, "neural": 26, "compressor": 26, "amd": 26, "viti": 26, "ai": 26, "autogptq": 27, "autoawq": 27, "workflow": [30, 37], "pip": [31, 32, 34], "sourc": 31, "edit": 31, "bert": 32, "cpu": [32, 34], "requir": [32, 34], "sampl": [32, 34], "The": 32, "techniqu": 32, "specifi": 32, "bert_ptq_cpu": 32, "automat": 32, "without": 32, "select": 32, "ani": 32, "quickstart": 33, "resnet": 34, "prepar": 34, "welcom": 35, "": 35, "document": 35, "overview": 35, "get": 35, "start": 35, "featur": 35, "extend": 35, "tutori": 35, "api": 35, "content": 36, "id": 37, "azureml_cli": 37, "aml_config_path": 37, "aml_config": 37, "field": 37, "inform": [37, 38], "flow": 37, "quick": 38, "tour": [38, 39], "need": 38, "acceler": [38, 47], "appli": 38, "advanc": 39, "self": 40, "kubernet": 40, "cluster": 40, "creat": 40, "extens": 40, "link": 40, "list": 41, "manage_compute_inst": 41, "usag": 41, "To": [42, 43, 44, 45, 46, 47], "auto": 42, "support": 43, "templat": 43, "titl": 43, "built": 43, "compon": 43, "type": 44, "accuraci": 44, "latenc": 44, "throughput": 44, "multi": 44, "set": 45, "path": 45, "folder": 45, "string": 45, "registri": 45, "nativ": 47, "readymad": 47, "isol": 47}, "envversion": {"sphinx.domains.c": 3, "sphinx.domains.changeset": 1, "sphinx.domains.citation": 1, "sphinx.domains.cpp": 9, "sphinx.domains.index": 1, "sphinx.domains.javascript": 3, "sphinx.domains.math": 2, "sphinx.domains.python": 4, "sphinx.domains.rst": 2, "sphinx.domains.std": 2, "sphinx.ext.intersphinx": 1, "sphinx.ext.todo": 2, "sphinx.ext.viewcode": 1, "sphinx": 58}, "alltitles": {"Engine": [[0, "engine"], [0, "id2"], [10, "engine"], [39, "engine"]], "EngineConfig": [[0, "engineconfig"]], "SearchStrategy": [[0, "searchstrategy"]], "OliveEvaluator": [[1, "oliveevaluator"], [1, "id1"]], "Metric": [[2, "metric"], [2, "id2"]], "MetricType": [[2, "metrictype"]], "AccuracySubType": [[2, "accuracysubtype"]], "LatencySubType": [[2, "latencysubtype"]], "ThroughputSubType": [[2, "throughputsubtype"]], "MetricGoal": [[2, "metricgoal"]], "OliveModels": [[3, "olivemodels"]], "Model Configuration": [[3, "model-configuration"]], "Hf Model Handler": [[3, "hf-model-handler"]], "Distributed Hf Model Handler": [[3, "distributed-hf-model-handler"]], "PyTorch Model Handler": [[3, "pytorch-model-handler"]], "ONNX Model Handler": [[3, "onnx-model-handler"]], "Distributed Onnx Model Handler": [[3, "distributed-onnx-model-handler"]], "OpenVINO Model Handler": [[3, "openvino-model-handler"]], "SNPE Model Handler": [[3, "snpe-model-handler"]], "Composite Model Handler": [[3, "composite-model-handler"]], "Passes": [[4, "passes"]], "OnnxConversion": [[4, "onnxconversion"]], "OnnxOpVersionConversion": [[4, "onnxopversionconversion"]], "OnnxModelOptimizer": [[4, "onnxmodeloptimizer"]], "OrtTransformersOptimization": [[4, "orttransformersoptimization"]], "OrtPerfTuning": [[4, "ortperftuning"]], "OnnxFloatToFloat16": [[4, "onnxfloattofloat16"]], "OnnxIOFloat16ToFloat32": [[4, "onnxiofloat16tofloat32"]], "OrtMixedPrecision": [[4, "ortmixedprecision"]], "QNNPreprocess": [[4, "qnnpreprocess"]], "MixedPrecisionOverrides": [[4, "mixedprecisionoverrides"]], "OnnxDynamicQuantization": [[4, "onnxdynamicquantization"]], "OnnxStaticQuantization": [[4, "onnxstaticquantization"]], "OnnxQuantization": [[4, "onnxquantization"]], "DynamicToFixedShape": [[4, "dynamictofixedshape"]], "IncDynamicQuantization": [[4, "incdynamicquantization"]], "IncStaticQuantization": [[4, "incstaticquantization"]], "IncQuantization": [[4, "incquantization"]], "VitisAIQuantization": [[4, "vitisaiquantization"]], "AppendPrePostProcessingOps": [[4, "appendprepostprocessingops"]], "InsertBeamSearch": [[4, "insertbeamsearch"]], "ExtractAdapters": [[4, "extractadapters"]], "LoRA": [[4, "lora"], [24, "lora"]], "QLoRA": [[4, "qlora"], [24, "qlora"]], "LoftQ": [[4, "loftq"], [24, "loftq"]], "LoRA/QLoRA/LoftQ HFTrainingArguments": [[4, "lora-qlora-loftq-hftrainingarguments"]], "QuantizationAwareTraining": [[4, "quantizationawaretraining"]], "OpenVINOConversion": [[4, "openvinoconversion"]], "OpenVINOQuantization": [[4, "openvinoquantization"]], "SNPEConversion": [[4, "snpeconversion"]], "SNPEQuantization": [[4, "snpequantization"]], "SNPEtoONNXConversion": [[4, "snpetoonnxconversion"]], "QNNConversion": [[4, "qnnconversion"]], "QNNModelLibGenerator": [[4, "qnnmodellibgenerator"]], "QNNContextBinaryGenerator": [[4, "qnncontextbinarygenerator"]], "MergeAdapterWeights": [[4, "mergeadapterweights"], [24, "mergeadapterweights"]], "SparseGPT": [[4, "sparsegpt"], [24, "sparsegpt"]], "SliceGPT": [[4, "slicegpt"], [24, "slicegpt"]], "QuaRot": [[4, "quarot"], [27, "quarot"]], "GptqQuantizer": [[4, "gptqquantizer"]], "AutoAWQQuantizer": [[4, "autoawqquantizer"]], "TorchTRTConversion": [[4, "torchtrtconversion"], [21, "torchtrtconversion"]], "OptimumConversion": [[4, "optimumconversion"]], "OptimumMerging": [[4, "optimummerging"]], "ModelBuilder": [[4, "modelbuilder"]], "ResourcePath": [[5, "resourcepath"]], "ResourceType": [[5, "resourcetype"]], "LocalFile": [[5, "localfile"]], "LocalFolder": [[5, "localfolder"]], "StringName": [[5, "stringname"]], "AzureMLModel": [[5, "azuremlmodel"]], "AzureMLDatastore": [[5, "azuremldatastore"]], "AzureMLJobOutput": [[5, "azuremljoboutput"]], "SearchAlgorithms": [[6, "searchalgorithms"]], "ExhaustiveSearchAlgorithm": [[6, "exhaustivesearchalgorithm"]], "RandomSearchAlgorithm": [[6, "randomsearchalgorithm"]], "TPESearchAlgorithm": [[6, "tpesearchalgorithm"]], "OliveSystems": [[7, "olivesystems"]], "Config": [[7, "config"]], "SystemType": [[7, "systemtype"]], "Device": [[7, "device"]], "AcceleratorConfig": [[7, "acceleratorconfig"], [47, null]], "LocalTargetUserConfig": [[7, "localtargetuserconfig"]], "DockerTargetUserConfig": [[7, "dockertargetuserconfig"]], "AzureMLTargetUserConfig": [[7, "azuremltargetuserconfig"]], "PythonEnvironmentTargetUserConfig": [[7, "pythonenvironmenttargetuserconfig"]], "IsolatedORTTargetUserConfig": [[7, "isolatedorttargetuserconfig"]], "Classes": [[7, "classes"]], "LocalSystem": [[7, "localsystem"]], "AzureMLSystem": [[7, "azuremlsystem"]], "DockerSystem": [[7, "dockersystem"]], "PythonEnvironmentSystem": [[7, "pythonenvironmentsystem"]], "IsolatedORTSystem": [[7, "isolatedortsystem"]], "System Alias": [[7, "module-olive.systems.system_alias"]], "Examples": [[8, "examples"], [9, "examples"], [9, "id1"]], "Custom Scripts": [[9, "custom-scripts"]], "user_script": [[9, "user-script"]], "script_dir": [[9, "script-dir"]], "Design": [[10, "design"]], "Pass": [[10, "pass"]], "System": [[10, "system"]], "Evaluator": [[10, "evaluator"], [39, "evaluator"]], "Search Strategy": [[10, "search-strategy"]], "Execution Order": [[10, "execution-order"]], "Search Algorithm": [[10, "search-algorithm"]], "How to add new optimization Pass": [[11, "how-to-add-new-optimization-pass"]], "1. Define a new class": [[11, "define-a-new-class"]], "2. Define configuration": [[11, "define-configuration"]], "Example": [[11, "example"], [37, "example"], [37, "id1"], [37, "id2"], [37, "id3"], [37, "id4"], [37, "id5"], [37, "id6"]], "3. Implement the run function": [[11, "implement-the-run-function"]], "Azure ML integration": [[12, "azure-ml-integration"]], "Azure Machine Learning client": [[12, "azure-machine-learning-client"]], "Olive Core": [[12, "olive-core"]], "Using AzureML registered model": [[12, "using-azureml-registered-model"]], "Using AzureML curated model": [[12, "using-azureml-curated-model"]], "Using model stored in AzureML datastore": [[12, "using-model-stored-in-azureml-datastore"]], "Using a model from an AzureML job output": [[12, "using-a-model-from-an-azureml-job-output"]], "Using data stored in AzureML datastore": [[12, "using-data-stored-in-azureml-datastore"]], "Using Azure ML compute as host or target": [[12, "using-azure-ml-compute-as-host-or-target"]], "Packaging the output models to AzureML workspace": [[12, "packaging-the-output-models-to-azureml-workspace"]], "Azure ML Models": [[12, "azure-ml-models"]], "Azure ML Data": [[12, "azure-ml-data"]], "Connect your own device to Azure ML as target or host by Azure Arc": [[12, "connect-your-own-device-to-azure-ml-as-target-or-host-by-azure-arc"]], "Azure ML Helper Scripts": [[12, "azure-ml-helper-scripts"]], "Command Line Tools": [[13, "command-line-tools"]], "Sub-commands": [[13, "Sub-commands"]], "run": [[13, "run"]], "Named Arguments": [[13, "named-arguments"], [13, "named-arguments_repeat1"], [13, "named-arguments_repeat2"], [13, "named-arguments_repeat3"], [13, "named-arguments_repeat4"], [13, "named-arguments_repeat5"], [13, "named-arguments_repeat6"]], "finetune": [[13, "finetune"]], "model options": [[13, "model-options"], [13, "model-options_repeat1"]], "dataset options": [[13, "dataset-options"]], "lora options": [[13, "lora-options"]], "remote options": [[13, "remote-options"], [13, "remote-options_repeat1"]], "export-adapters": [[13, "export-adapters"]], "int4 quantization options": [[13, "int4-quantization-options"]], "configure-qualcomm-sdk": [[13, "configure-qualcomm-sdk"]], "manage-aml-compute": [[13, "manage-aml-compute"]], "tune-session-params": [[13, "tune-session-params"]], "dataset options, which mutually exclusive with huggingface dataset options": [[13, "dataset-options,-which-mutually-exclusive-with-huggingface-dataset-options"]], "huggingface dataset options, if dataset options are not provided, user should provide the following options to modify the default data config. Please refer to olive.data.container.TransformersTokenDummyDataContainer for more details.": [[13, "huggingface-dataset-options,-if-dataset-options-are-not-provided,-user-should-provide-the-following-options-to-modify-the-default-data-config.-please-refer-to-olive.data.container.transformerstokendummydatacontainer-for-more-details."]], "pass options": [[13, "pass-options"]], "cloud-cache": [[13, "cloud-cache"]], "Cloud Model Cache": [[14, "cloud-model-cache"]], "What is Cloud Model Cache": [[14, "what-is-cloud-model-cache"]], "How to use Cloud Model Cache": [[14, "how-to-use-cloud-model-cache"]], "Install Extra Dependencies": [[14, "install-extra-dependencies"], [30, "install-extra-dependencies"]], "Azure Blob Storage": [[14, "azure-blob-storage"]], "Cloud Cache Configuration": [[14, "cloud-cache-configuration"]], "Notes": [[14, "notes"]], "Model Conversions": [[15, "model-conversions"]], "Huggingface Integration": [[16, "huggingface-integration"]], "Introduction": [[16, "introduction"]], "Input Model": [[16, "input-model"], [38, "input-model"], [39, "input-model"]], "Huggingface Hub model": [[16, "huggingface-hub-model"]], "Local model": [[16, "local-model"]], "Azure ML model": [[16, "azure-ml-model"]], "Model config loading": [[16, "model-config-loading"]], "Huggingface datasets": [[16, "huggingface-datasets"]], "Huggingface metrics": [[16, "huggingface-metrics"]], "Huggingface login": [[16, "huggingface-login"]], "Local system, docker system and Python environment system": [[16, "local-system-docker-system-and-python-environment-system"]], "AzureML system": [[16, "azureml-system"]], "E2E example": [[16, "e2e-example"]], "LoRA Adapters": [[17, "lora-adapters"]], "Model Transformations and Optimizations": [[18, "model-transformations-and-optimizations"]], "Packaging Olive artifacts": [[19, "packaging-olive-artifacts"]], "What is Olive Packaging": [[19, "what-is-olive-packaging"]], "Zipfile": [[19, "zipfile"]], "CandidateModels": [[19, "candidatemodels"]], "Models rank JSON file": [[19, "models-rank-json-file"]], "AzureMLModels": [[19, "azuremlmodels"]], "AzureMLData": [[19, "azuremldata"]], "AzureMLDeployment": [[19, "azuremldeployment"]], "Dockerfile": [[19, "dockerfile"]], "How to package Olive artifacts": [[19, "how-to-package-olive-artifacts"]], "Packaged files": [[19, "packaged-files"]], "Inference config file": [[19, "inference-config-file"]], "Model configuration file": [[19, "model-configuration-file"]], "Metrics file": [[19, "metrics-file"]], "ONNX": [[20, "onnx"], [22, "onnx"], [26, "onnx"]], "Model Conversion": [[20, "model-conversion"], [23, "model-conversion"], [28, "model-conversion"]], "Example Configuration": [[20, "example-configuration"], [20, "id1"], [20, "id2"], [20, "id3"], [20, "id4"], [20, "id5"], [21, "example-configuration"], [22, "example-configuration"], [22, "id1"], [22, "id2"], [22, "id3"], [22, "id4"], [22, "id5"], [23, "example-configuration"], [23, "id1"], [24, "example-configuration"], [24, "id1"], [24, "id2"], [24, "id3"], [24, "id4"], [24, "id5"], [25, "example-configuration"], [25, "id1"], [25, "id2"], [26, "example-configuration"], [26, "id1"], [26, "id2"], [27, "example-configuration"], [27, "id1"], [27, "id2"], [28, "example-configuration"], [28, "id1"]], "Float16 Conversion": [[20, "float16-conversion"]], "Inputs/Outputs Float16 to Float32 Conversion": [[20, "inputs-outputs-float16-to-float32-conversion"]], "Mixed Precision Conversion": [[20, "mixed-precision-conversion"]], "Convert dynamic shape to fixed shape": [[20, "convert-dynamic-shape-to-fixed-shape"]], "PyTorch": [[21, "pytorch"], [24, "pytorch"], [27, "pytorch"]], "Model Optimizer": [[22, "model-optimizer"]], "ORT Transformers Optimization": [[22, "ort-transformers-optimization"]], "Append Pre/Post Processing Ops": [[22, "append-pre-post-processing-ops"]], "Insert Beam Search Op": [[22, "insert-beam-search-op"]], "ORT Performance Tuning": [[22, "ort-performance-tuning"]], "Extract Adapters": [[22, "extract-adapters"]], "OpenVINO": [[23, "openvino"]], "Prerequisites": [[23, "prerequisites"], [25, "prerequisites"], [28, "prerequisites"], [32, "prerequisites"], [34, "prerequisites"], [47, "prerequisites"], [47, "id2"]], "Option 1: install Olive with OpenVINO extras": [[23, "option-1-install-olive-with-openvino-extras"]], "Option 2: Install OpenVINO Runtime and OpenVINO Development Tools from Pypi": [[23, "option-2-install-openvino-runtime-and-openvino-development-tools-from-pypi"]], "Post Training Quantization (PTQ)": [[23, "post-training-quantization-ptq"], [28, "post-training-quantization-ptq"]], "Quantization Aware Training": [[24, "quantization-aware-training"]], "QNN": [[25, "qnn"]], "Download and unzip QNN SDK": [[25, "download-and-unzip-qnn-sdk"]], "Configure Olive QNN": [[25, "configure-olive-qnn"]], "Model Conversion/Quantization": [[25, "model-conversion-quantization"]], "Model Library Generation": [[25, "model-library-generation"]], "Context Binary Generation": [[25, "context-binary-generation"]], "Quantize with onnxruntime": [[26, "quantize-with-onnxruntime"]], "Quantize with Intel\u00ae Neural Compressor": [[26, "quantize-with-intel-neural-compressor"]], "Quantize with AMD Vitis AI Quantizer": [[26, "quantize-with-amd-vitis-ai-quantizer"]], "AutoGPTQ": [[27, "autogptq"]], "AutoAWQ": [[27, "autoawq"]], "SNPE": [[28, "snpe"]], "Download and unzip SNPE SDK": [[28, "download-and-unzip-snpe-sdk"]], "Configure Olive SNPE": [[28, "configure-olive-snpe"]], "Model Quantizations": [[29, "model-quantizations"]], "Remote Workflow": [[30, "remote-workflow"]], "Running Olive Workflow Remotely on Azure Machine Learning workspace compute": [[30, "running-olive-workflow-remotely-on-azure-machine-learning-workspace-compute"]], "Configure an AzureML system": [[30, "configure-an-azureml-system"]], "Run Olive Workflow": [[30, "run-olive-workflow"]], "Workflow outputs": [[30, "workflow-outputs"]], "Installation": [[31, "installation"]], "Install with pip": [[31, "install-with-pip"]], "Install from source": [[31, "install-from-source"]], "Editable install": [[31, "editable-install"]], "Optional Dependencies": [[31, "optional-dependencies"]], "BERT optimization with PTQ on CPU": [[32, "bert-optimization-with-ptq-on-cpu"]], "Pip requirements": [[32, "pip-requirements"], [34, "pip-requirements"]], "Run sample using config. The optimization techniques to run are specified in bert_ptq_cpu.json": [[32, "run-sample-using-config-the-optimization-techniques-to-run-are-specified-in-bert-ptq-cpu-json"]], "Optimize model automatically without selecting any optimization technique.": [[32, "optimize-model-automatically-without-selecting-any-optimization-technique"]], "Quickstart Examples": [[33, "quickstart-examples"]], "ResNet optimization with PTQ on CPU": [[34, "resnet-optimization-with-ptq-on-cpu"]], "Prepare data and model": [[34, "prepare-data-and-model"]], "Run sample using config": [[34, "run-sample-using-config"]], "Welcome to Olive\u2019s documentation!": [[35, "welcome-to-olive-s-documentation"]], "OVERVIEW": [[35, null]], "GET STARTED": [[35, null]], "EXAMPLES": [[35, null]], "FEATURES": [[35, null]], "EXTENDING OLIVE": [[35, null]], "TUTORIALS": [[35, null]], "API REFERENCE": [[35, null]], "Olive": [[36, "olive"]], "Contents": [[36, "contents"]], "Olive Options": [[37, "olive-options"]], "Workflow ID": [[37, "workflow-id"]], "Workflow Host": [[37, "workflow-host"]], "Azure ML Client": [[37, "azure-ml-client"]], "azureml_client with aml_config_path:": [[37, "azureml-client-with-aml-config-path"]], "aml_config.json:": [[37, "aml-config-json"]], "azureml_client:": [[37, "azureml-client"]], "azureml_client with azureml config fields:": [[37, "azureml-client-with-azureml-config-fields"]], "Input Model Information": [[37, "input-model-information"]], "Systems Information": [[37, "systems-information"]], "Evaluators Information": [[37, "evaluators-information"]], "Passes Information": [[37, "passes-information"]], "Pass Flows Information": [[37, "pass-flows-information"]], "Engine Information": [[37, "engine-information"]], "Quick Tour": [[38, "quick-tour"]], "Information needed to accelerate a model": [[38, "information-needed-to-accelerate-a-model"]], "Passes to apply": [[38, "passes-to-apply"]], "Olive Engine": [[38, "olive-engine"]], "Advanced User Tour": [[39, "advanced-user-tour"]], "Host and Target Systems": [[39, "host-and-target-systems"]], "Register Passes": [[39, "register-passes"]], "Run the engine": [[39, "run-the-engine"]], "Self-hosted Kubernetes cluster": [[40, "self-hosted-kubernetes-cluster"]], "Create Kubernetes cluster": [[40, "create-kubernetes-cluster"]], "Add local Kubernetes cluster to Azure Arc": [[40, "add-local-kubernetes-cluster-to-azure-arc"]], "Add AML extension to cluster": [[40, "add-aml-extension-to-cluster"]], "Link Azure Arc Kubernetes cluster to Azure Machine Learning": [[40, "link-azure-arc-kubernetes-cluster-to-azure-machine-learning"]], "Azure ML scripts": [[41, "azure-ml-scripts"]], "Scripts list": [[41, "scripts-list"]], "manage_compute_instance": [[41, "manage-compute-instance"]], "Usage": [[41, "usage"]], "How To Configure Auto Optimizer": [[42, "how-to-configure-auto-optimizer"]], "Auto Optimizer Configuration": [[42, "auto-optimizer-configuration"]], "How To Configure Data": [[43, "how-to-configure-data"]], "Supported Data Config Template": [[43, "supported-data-config-template"]], "Generic Data Config": [[43, "generic-data-config"]], "Title": [[43, "id2"]], "Configs with built-in component:": [[43, "configs-with-built-in-component"]], "Configs with customized component:": [[43, "configs-with-customized-component"]], "How To Configure Metric": [[44, "how-to-configure-metric"]], "Metric Types": [[44, "metric-types"]], "Accuracy Metric": [[44, "accuracy-metric"]], "Latency Metric": [[44, "latency-metric"]], "Throughput Metric": [[44, "throughput-metric"]], "Custom Metric": [[44, "custom-metric"]], "Multi Metrics configuration": [[44, "multi-metrics-configuration"]], "How To Set Model Path": [[45, "how-to-set-model-path"]], "Local Model Path": [[45, "local-model-path"]], "Local File": [[45, "local-file"]], "Local Folder": [[45, "local-folder"]], "String Name": [[45, "string-name"]], "Remote Model Path": [[45, "remote-model-path"]], "AzureML Model": [[45, "azureml-model"]], "AzureML Registry Model": [[45, "azureml-registry-model"]], "AzureML Datastore": [[45, "azureml-datastore"]], "AzureML Job Output": [[45, "azureml-job-output"]], "How To Configure Pass": [[46, "how-to-configure-pass"]], "How To Configure System": [[47, "how-to-configure-system"]], "Accelerator Configuration": [[47, "accelerator-configuration"]], "Local System": [[47, "local-system"]], "Configuration": [[47, null], [47, null], [47, null], [47, null], [47, null]], "Python Environment System": [[47, "python-environment-system"]], "Native Python Environment System": [[47, "native-python-environment-system"]], "Managed Python Environment System": [[47, "managed-python-environment-system"]], "Docker System": [[47, "docker-system"]], "Native Docker System": [[47, "native-docker-system"]], "Managed Docker System": [[47, "managed-docker-system"]], "AzureML System": [[47, "azureml-system"]], "Native AzureML System": [[47, "native-azureml-system"]], "AzureML Readymade Systems": [[47, "azureml-readymade-systems"]], "Managed AzureML System": [[47, "managed-azureml-system"]], "Isolated ORT System": [[47, "isolated-ort-system"]]}, "indexentries": {"engine (class in olive.engine)": [[0, "olive.engine.Engine"]], "searchstrategy (class in olive.strategy.search_strategy)": [[0, "olive.strategy.search_strategy.SearchStrategy"]], "cache_dir (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.cache_dir"]], "clean_cache (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.clean_cache"]], "clean_evaluation_cache (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.clean_evaluation_cache"]], "create_evaluator() (olive.evaluator.olive_evaluator.oliveevaluatorconfig method)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.create_evaluator"]], "evaluator (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.evaluator"]], "execution_order (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.execution_order"]], "host (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.host"]], "is_accuracy_drop_tolerance (olive.evaluator.olive_evaluator.oliveevaluatorconfig property)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.is_accuracy_drop_tolerance"]], "max_iter (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.max_iter"]], "max_time (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.max_time"]], "metrics (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.metrics"]], "output_model_num (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.output_model_num"]], "plot_pareto_frontier (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.plot_pareto_frontier"]], "register() (olive.engine.engine method)": [[0, "olive.engine.Engine.register"]], "run() (olive.engine.engine method)": [[0, "olive.engine.Engine.run"]], "script_dir (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.script_dir"]], "search_algorithm (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.search_algorithm"]], "search_algorithm_config (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.search_algorithm_config"]], "search_strategy (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.search_strategy"]], "stop_when_goals_met (olive.strategy.search_strategy.searchstrategyconfig attribute)": [[0, "olive.strategy.search_strategy.SearchStrategyConfig.stop_when_goals_met"]], "target (olive.engine.engineconfig attribute)": [[0, "olive.engine.EngineConfig.target"]], "type (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.type"]], "type_args (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.type_args"]], "user_script (olive.evaluator.olive_evaluator.oliveevaluatorconfig attribute)": [[0, "olive.evaluator.olive_evaluator.OliveEvaluatorConfig.user_script"]], "oliveevaluator (class in olive.evaluator.olive_evaluator)": [[1, "olive.evaluator.olive_evaluator.OliveEvaluator"]], "accuracy (olive.evaluator.metric.metrictype attribute)": [[2, "olive.evaluator.metric.MetricType.ACCURACY"]], "accuracy_score (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.ACCURACY_SCORE"]], "auroc (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.AUROC"]], "avg (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.AVG"]], "avg (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.AVG"]], "accuracysubtype (class in olive.evaluator.metric)": [[2, "olive.evaluator.metric.AccuracySubType"]], "custom (olive.evaluator.metric.metrictype attribute)": [[2, "olive.evaluator.metric.MetricType.CUSTOM"]], "f1_score (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.F1_SCORE"]], "latency (olive.evaluator.metric.metrictype attribute)": [[2, "olive.evaluator.metric.MetricType.LATENCY"]], "latencysubtype (class in olive.evaluator.metric)": [[2, "olive.evaluator.metric.LatencySubType"]], "max (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.MAX"]], "max (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.MAX"]], "min (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.MIN"]], "min (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.MIN"]], "metrictype (class in olive.evaluator.metric)": [[2, "olive.evaluator.metric.MetricType"]], "p50 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P50"]], "p50 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P50"]], "p75 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P75"]], "p75 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P75"]], "p90 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P90"]], "p90 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P90"]], "p95 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P95"]], "p95 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P95"]], "p99 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P99"]], "p99 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P99"]], "p999 (olive.evaluator.metric.latencysubtype attribute)": [[2, "olive.evaluator.metric.LatencySubType.P999"]], "p999 (olive.evaluator.metric.throughputsubtype attribute)": [[2, "olive.evaluator.metric.ThroughputSubType.P999"]], "perplexity (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.PERPLEXITY"]], "precision (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.PRECISION"]], "recall (olive.evaluator.metric.accuracysubtype attribute)": [[2, "olive.evaluator.metric.AccuracySubType.RECALL"]], "throughput (olive.evaluator.metric.metrictype attribute)": [[2, "olive.evaluator.metric.MetricType.THROUGHPUT"]], "throughputsubtype (class in olive.evaluator.metric)": [[2, "olive.evaluator.metric.ThroughputSubType"]], "backend (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.backend"]], "data_config (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.data_config"]], "get_inference_settings() (olive.evaluator.metric.metric method)": [[2, "olive.evaluator.metric.Metric.get_inference_settings"]], "get_run_kwargs() (olive.evaluator.metric.metric method)": [[2, "olive.evaluator.metric.Metric.get_run_kwargs"]], "get_sub_type_info() (olive.evaluator.metric.metric method)": [[2, "olive.evaluator.metric.Metric.get_sub_type_info"]], "has_regression_goal() (olive.evaluator.metric.metricgoal method)": [[2, "olive.evaluator.metric.MetricGoal.has_regression_goal"]], "name (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.name"]], "sub_types (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.sub_types"]], "type (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.type"]], "type (olive.evaluator.metric.metricgoal attribute)": [[2, "olive.evaluator.metric.MetricGoal.type"]], "user_config (olive.evaluator.metric.metric attribute)": [[2, "olive.evaluator.metric.Metric.user_config"]], "value (olive.evaluator.metric.metricgoal attribute)": [[2, "olive.evaluator.metric.MetricGoal.value"]], "compositemodelhandler (class in olive.model)": [[3, "olive.model.CompositeModelHandler"]], "distributedhfmodelhandler (class in olive.model)": [[3, "olive.model.DistributedHfModelHandler"]], "distributedonnxmodelhandler (class in olive.model)": [[3, "olive.model.DistributedOnnxModelHandler"]], "hfmodelhandler (class in olive.model)": [[3, "olive.model.HfModelHandler"]], "modelconfig (class in olive.model)": [[3, "olive.model.ModelConfig"]], "onnxmodelhandler (class in olive.model)": [[3, "olive.model.ONNXModelHandler"]], "openvinomodelhandler (class in olive.model)": [[3, "olive.model.OpenVINOModelHandler"]], "pytorchmodelhandler (class in olive.model)": [[3, "olive.model.PyTorchModelHandler"]], "snpemodelhandler (class in olive.model)": [[3, "olive.model.SNPEModelHandler"]], "activationsymmetric": [[4, "cmdoption-arg-164"], [4, "cmdoption-arg-62"], [4, "cmdoption-arg-89"], [4, "cmdoption-arg-ActivationSymmetric"]], "addqdqpairtoweight": [[4, "cmdoption-arg-AddQDQPairToWeight"]], "enablesubgraph": [[4, "cmdoption-arg-64"], [4, "cmdoption-arg-91"], [4, "cmdoption-arg-EnableSubgraph"]], "forcequantizenoinputcheck": [[4, "cmdoption-arg-65"], [4, "cmdoption-arg-92"], [4, "cmdoption-arg-ForceQuantizeNoInputCheck"]], "matmulconstbonly": [[4, "cmdoption-arg-66"], [4, "cmdoption-arg-93"], [4, "cmdoption-arg-MatMulConstBOnly"]], "weightsymmetric": [[4, "cmdoption-arg-165"], [4, "cmdoption-arg-63"], [4, "cmdoption-arg-90"], [4, "cmdoption-arg-WeightSymmetric"]], "a_asym": [[4, "cmdoption-arg-a_asym"]], "a_bits": [[4, "cmdoption-arg-a_bits"]], "a_clip_ratio": [[4, "cmdoption-arg-a_clip_ratio"]], "a_groupsize": [[4, "cmdoption-arg-a_groupsize"]], "activation_type": [[4, "cmdoption-arg-163"], [4, "cmdoption-arg-85"], [4, "cmdoption-arg-activation_type"]], "all_tensors_to_one_file": [[4, "cmdoption-arg-101"], [4, "cmdoption-arg-108"], [4, "cmdoption-arg-12"], [4, "cmdoption-arg-128"], [4, "cmdoption-arg-149"], [4, "cmdoption-arg-168"], [4, "cmdoption-arg-174"], [4, "cmdoption-arg-180"], [4, "cmdoption-arg-185"], [4, "cmdoption-arg-2"], [4, "cmdoption-arg-21"], [4, "cmdoption-arg-224"], [4, "cmdoption-arg-259"], [4, "cmdoption-arg-26"], [4, "cmdoption-arg-32"], [4, "cmdoption-arg-37"], [4, "cmdoption-arg-42"], [4, "cmdoption-arg-47"], [4, "cmdoption-arg-69"], [4, "cmdoption-arg-7"], [4, "cmdoption-arg-96"], [4, "cmdoption-arg-all_tensors_to_one_file"]], "allow_tf32": [[4, "cmdoption-arg-195"], [4, "cmdoption-arg-205"], [4, "cmdoption-arg-allow_tf32"]], "append_first_op_types_to_quantize_list": [[4, "cmdoption-arg-54"], [4, "cmdoption-arg-76"], [4, "cmdoption-arg-append_first_op_types_to_quantize_list"]], "approach": [[4, "cmdoption-arg-112"], [4, "cmdoption-arg-132"], [4, "cmdoption-arg-approach"]], "atol": [[4, "cmdoption-arg-atol"]], "backend": [[4, "cmdoption-arg-114"], [4, "cmdoption-arg-134"], [4, "cmdoption-arg-229"], [4, "cmdoption-arg-backend"]], "binary_file": [[4, "cmdoption-arg-binary_file"]], "bits": [[4, "cmdoption-arg-bits"]], "blocksize": [[4, "cmdoption-arg-blocksize"]], "calibrate_method": [[4, "cmdoption-arg-161"], [4, "cmdoption-arg-83"], [4, "cmdoption-arg-calibrate_method"]], "calibration_batch_size": [[4, "cmdoption-arg-237"], [4, "cmdoption-arg-calibration_batch_size"]], "calibration_data_config": [[4, "cmdoption-arg-235"], [4, "cmdoption-arg-calibration_data_config"]], "calibration_nsamples": [[4, "cmdoption-arg-236"], [4, "cmdoption-arg-calibration_nsamples"]], "calibration_sampling_size": [[4, "cmdoption-arg-147"], [4, "cmdoption-arg-calibration_sampling_size"]], "checkpoint_path": [[4, "cmdoption-arg-checkpoint_path"]], "command line option": [[4, "cmdoption-arg-0"], [4, "cmdoption-arg-1"], [4, "cmdoption-arg-10"], [4, "cmdoption-arg-100"], [4, "cmdoption-arg-101"], [4, "cmdoption-arg-102"], [4, "cmdoption-arg-103"], [4, "cmdoption-arg-104"], [4, "cmdoption-arg-105"], [4, "cmdoption-arg-106"], [4, "cmdoption-arg-107"], [4, "cmdoption-arg-108"], [4, "cmdoption-arg-109"], [4, "cmdoption-arg-11"], [4, "cmdoption-arg-110"], [4, "cmdoption-arg-111"], [4, "cmdoption-arg-112"], [4, "cmdoption-arg-113"], [4, "cmdoption-arg-114"], [4, "cmdoption-arg-115"], [4, "cmdoption-arg-116"], [4, "cmdoption-arg-117"], [4, "cmdoption-arg-118"], [4, "cmdoption-arg-119"], [4, "cmdoption-arg-12"], [4, "cmdoption-arg-120"], [4, "cmdoption-arg-121"], [4, "cmdoption-arg-122"], [4, "cmdoption-arg-123"], [4, "cmdoption-arg-124"], [4, "cmdoption-arg-125"], [4, "cmdoption-arg-126"], [4, "cmdoption-arg-127"], [4, "cmdoption-arg-128"], [4, "cmdoption-arg-129"], [4, "cmdoption-arg-13"], [4, "cmdoption-arg-130"], [4, "cmdoption-arg-131"], [4, "cmdoption-arg-132"], [4, "cmdoption-arg-133"], [4, "cmdoption-arg-134"], [4, "cmdoption-arg-135"], [4, "cmdoption-arg-136"], [4, "cmdoption-arg-137"], [4, "cmdoption-arg-138"], [4, "cmdoption-arg-139"], [4, "cmdoption-arg-14"], [4, "cmdoption-arg-140"], [4, "cmdoption-arg-141"], [4, "cmdoption-arg-142"], [4, "cmdoption-arg-143"], [4, "cmdoption-arg-144"], [4, "cmdoption-arg-145"], [4, "cmdoption-arg-146"], [4, "cmdoption-arg-147"], [4, "cmdoption-arg-148"], [4, "cmdoption-arg-149"], [4, "cmdoption-arg-15"], [4, "cmdoption-arg-150"], [4, "cmdoption-arg-151"], [4, "cmdoption-arg-152"], [4, "cmdoption-arg-153"], [4, "cmdoption-arg-154"], [4, "cmdoption-arg-155"], [4, "cmdoption-arg-156"], [4, "cmdoption-arg-157"], [4, "cmdoption-arg-158"], [4, "cmdoption-arg-159"], [4, "cmdoption-arg-16"], [4, "cmdoption-arg-160"], [4, "cmdoption-arg-161"], [4, "cmdoption-arg-162"], [4, "cmdoption-arg-163"], [4, "cmdoption-arg-164"], [4, "cmdoption-arg-165"], [4, "cmdoption-arg-166"], [4, "cmdoption-arg-167"], [4, "cmdoption-arg-168"], [4, "cmdoption-arg-169"], [4, "cmdoption-arg-17"], [4, "cmdoption-arg-170"], [4, "cmdoption-arg-171"], [4, "cmdoption-arg-172"], [4, "cmdoption-arg-173"], [4, "cmdoption-arg-174"], [4, "cmdoption-arg-175"], [4, "cmdoption-arg-176"], [4, "cmdoption-arg-177"], [4, "cmdoption-arg-178"], [4, "cmdoption-arg-179"], [4, "cmdoption-arg-18"], [4, "cmdoption-arg-180"], [4, "cmdoption-arg-181"], [4, "cmdoption-arg-182"], [4, "cmdoption-arg-183"], [4, "cmdoption-arg-184"], [4, "cmdoption-arg-185"], [4, "cmdoption-arg-186"], [4, "cmdoption-arg-187"], [4, "cmdoption-arg-188"], [4, "cmdoption-arg-189"], [4, "cmdoption-arg-19"], [4, "cmdoption-arg-190"], [4, "cmdoption-arg-191"], [4, "cmdoption-arg-192"], [4, "cmdoption-arg-193"], [4, "cmdoption-arg-194"], [4, "cmdoption-arg-195"], [4, "cmdoption-arg-196"], [4, "cmdoption-arg-197"], [4, "cmdoption-arg-198"], [4, "cmdoption-arg-199"], [4, "cmdoption-arg-2"], [4, "cmdoption-arg-20"], [4, "cmdoption-arg-200"], [4, "cmdoption-arg-201"], [4, "cmdoption-arg-202"], [4, "cmdoption-arg-203"], [4, "cmdoption-arg-204"], [4, "cmdoption-arg-205"], [4, "cmdoption-arg-206"], [4, "cmdoption-arg-207"], [4, "cmdoption-arg-208"], [4, "cmdoption-arg-209"], [4, "cmdoption-arg-21"], [4, "cmdoption-arg-210"], [4, "cmdoption-arg-211"], [4, "cmdoption-arg-212"], [4, "cmdoption-arg-213"], [4, "cmdoption-arg-214"], [4, "cmdoption-arg-215"], [4, "cmdoption-arg-216"], [4, "cmdoption-arg-217"], [4, "cmdoption-arg-218"], [4, "cmdoption-arg-219"], [4, "cmdoption-arg-22"], [4, "cmdoption-arg-220"], [4, "cmdoption-arg-221"], [4, "cmdoption-arg-222"], [4, "cmdoption-arg-223"], [4, "cmdoption-arg-224"], [4, "cmdoption-arg-225"], [4, "cmdoption-arg-226"], [4, "cmdoption-arg-227"], [4, "cmdoption-arg-228"], [4, "cmdoption-arg-229"], [4, "cmdoption-arg-23"], [4, "cmdoption-arg-230"], [4, "cmdoption-arg-231"], [4, "cmdoption-arg-232"], [4, "cmdoption-arg-233"], [4, "cmdoption-arg-234"], [4, "cmdoption-arg-235"], [4, "cmdoption-arg-236"], [4, "cmdoption-arg-237"], [4, "cmdoption-arg-238"], [4, "cmdoption-arg-239"], [4, "cmdoption-arg-24"], [4, "cmdoption-arg-240"], [4, "cmdoption-arg-241"], [4, "cmdoption-arg-242"], [4, "cmdoption-arg-243"], [4, "cmdoption-arg-244"], [4, "cmdoption-arg-245"], [4, "cmdoption-arg-246"], [4, "cmdoption-arg-247"], [4, "cmdoption-arg-248"], [4, "cmdoption-arg-249"], [4, "cmdoption-arg-25"], [4, "cmdoption-arg-250"], [4, "cmdoption-arg-251"], [4, "cmdoption-arg-252"], [4, "cmdoption-arg-253"], [4, "cmdoption-arg-254"], [4, "cmdoption-arg-255"], [4, "cmdoption-arg-256"], [4, "cmdoption-arg-257"], [4, "cmdoption-arg-258"], [4, "cmdoption-arg-259"], [4, "cmdoption-arg-26"], [4, "cmdoption-arg-260"], [4, "cmdoption-arg-261"], [4, "cmdoption-arg-262"], [4, "cmdoption-arg-263"], [4, "cmdoption-arg-27"], [4, "cmdoption-arg-28"], [4, "cmdoption-arg-29"], [4, "cmdoption-arg-3"], [4, "cmdoption-arg-30"], [4, "cmdoption-arg-31"], [4, "cmdoption-arg-32"], [4, "cmdoption-arg-33"], [4, "cmdoption-arg-34"], [4, "cmdoption-arg-35"], [4, "cmdoption-arg-36"], [4, "cmdoption-arg-37"], [4, "cmdoption-arg-38"], [4, "cmdoption-arg-39"], [4, "cmdoption-arg-4"], [4, "cmdoption-arg-40"], [4, "cmdoption-arg-41"], [4, "cmdoption-arg-42"], [4, "cmdoption-arg-43"], [4, "cmdoption-arg-44"], [4, "cmdoption-arg-45"], [4, "cmdoption-arg-46"], [4, "cmdoption-arg-47"], [4, "cmdoption-arg-48"], [4, "cmdoption-arg-49"], [4, "cmdoption-arg-5"], [4, "cmdoption-arg-50"], [4, "cmdoption-arg-51"], [4, "cmdoption-arg-52"], [4, "cmdoption-arg-53"], [4, "cmdoption-arg-54"], [4, "cmdoption-arg-55"], [4, "cmdoption-arg-56"], [4, "cmdoption-arg-57"], [4, "cmdoption-arg-58"], [4, "cmdoption-arg-59"], [4, "cmdoption-arg-6"], [4, "cmdoption-arg-60"], [4, "cmdoption-arg-61"], [4, "cmdoption-arg-62"], [4, "cmdoption-arg-63"], [4, "cmdoption-arg-64"], [4, "cmdoption-arg-65"], [4, "cmdoption-arg-66"], [4, "cmdoption-arg-67"], [4, "cmdoption-arg-68"], [4, "cmdoption-arg-69"], [4, "cmdoption-arg-7"], [4, "cmdoption-arg-70"], [4, "cmdoption-arg-71"], [4, "cmdoption-arg-72"], [4, "cmdoption-arg-73"], [4, "cmdoption-arg-74"], [4, "cmdoption-arg-75"], [4, "cmdoption-arg-76"], [4, "cmdoption-arg-77"], [4, "cmdoption-arg-78"], [4, "cmdoption-arg-79"], [4, "cmdoption-arg-8"], [4, "cmdoption-arg-80"], [4, "cmdoption-arg-81"], [4, "cmdoption-arg-82"], [4, "cmdoption-arg-83"], [4, "cmdoption-arg-84"], [4, "cmdoption-arg-85"], [4, "cmdoption-arg-86"], [4, "cmdoption-arg-87"], [4, "cmdoption-arg-88"], [4, "cmdoption-arg-89"], [4, "cmdoption-arg-9"], [4, "cmdoption-arg-90"], [4, "cmdoption-arg-91"], [4, "cmdoption-arg-92"], [4, "cmdoption-arg-93"], [4, "cmdoption-arg-94"], [4, "cmdoption-arg-95"], [4, "cmdoption-arg-96"], [4, "cmdoption-arg-97"], [4, "cmdoption-arg-98"], [4, "cmdoption-arg-99"], [4, "cmdoption-arg-ActivationSymmetric"], [4, "cmdoption-arg-AddQDQPairToWeight"], [4, "cmdoption-arg-EnableSubgraph"], [4, "cmdoption-arg-ForceQuantizeNoInputCheck"], [4, "cmdoption-arg-MatMulConstBOnly"], [4, "cmdoption-arg-WeightSymmetric"], [4, "cmdoption-arg-a_asym"], [4, "cmdoption-arg-a_bits"], [4, "cmdoption-arg-a_clip_ratio"], [4, "cmdoption-arg-a_groupsize"], [4, "cmdoption-arg-activation_type"], [4, "cmdoption-arg-all_tensors_to_one_file"], [4, "cmdoption-arg-allow_tf32"], [4, "cmdoption-arg-append_first_op_types_to_quantize_list"], [4, "cmdoption-arg-approach"], [4, "cmdoption-arg-atol"], [4, "cmdoption-arg-backend"], [4, "cmdoption-arg-binary_file"], [4, "cmdoption-arg-bits"], [4, "cmdoption-arg-blocksize"], [4, "cmdoption-arg-calibrate_method"], [4, "cmdoption-arg-calibration_batch_size"], [4, "cmdoption-arg-calibration_data_config"], [4, "cmdoption-arg-calibration_nsamples"], [4, "cmdoption-arg-calibration_sampling_size"], [4, "cmdoption-arg-checkpoint_path"], [4, "cmdoption-arg-components"], [4, "cmdoption-arg-compress_to_fp16"], [4, "cmdoption-arg-compute_dtype"], [4, "cmdoption-arg-convert_attribute"], [4, "cmdoption-arg-cpu_cores"], [4, "cmdoption-arg-damp_percent"], [4, "cmdoption-arg-data_config"], [4, "cmdoption-arg-desc_act"], [4, "cmdoption-arg-device"], [4, "cmdoption-arg-dim_param"], [4, "cmdoption-arg-dim_value"], [4, "cmdoption-arg-do_validate"], [4, "cmdoption-arg-domain"], [4, "cmdoption-arg-double_quant"], [4, "cmdoption-arg-duo_scaling"], [4, "cmdoption-arg-dynamic_lora_r"], [4, "cmdoption-arg-element_wise_binary_ops"], [4, "cmdoption-arg-enable_cuda_graph"], [4, "cmdoption-arg-enable_dpu"], [4, "cmdoption-arg-enable_htp"], [4, "cmdoption-arg-enable_profiling"], [4, "cmdoption-arg-eval_data_config"], [4, "cmdoption-arg-example_input_func"], [4, "cmdoption-arg-exclude_embeds"], [4, "cmdoption-arg-exclude_lm_head"], [4, "cmdoption-arg-excluded_precisions"], [4, "cmdoption-arg-execution_mode_list"], [4, "cmdoption-arg-export_compatible"], [4, "cmdoption-arg-external_data_name"], [4, "cmdoption-arg-extra.Sigmoid.nnapi"], [4, "cmdoption-arg-extra_args"], [4, "cmdoption-arg-extra_configs"], [4, "cmdoption-arg-extra_options"], [4, "cmdoption-arg-extra_session_config"], [4, "cmdoption-arg-final_orientation"], [4, "cmdoption-arg-float16"], [4, "cmdoption-arg-force_evaluate_other_eps"], [4, "cmdoption-arg-force_fp16_inputs"], [4, "cmdoption-arg-force_fp32_nodes"], [4, "cmdoption-arg-force_fp32_ops"], [4, "cmdoption-arg-fp16"], [4, "cmdoption-arg-fuse_layernorm"], [4, "cmdoption-arg-gptq_damping"], [4, "cmdoption-arg-gptq_opt_scales"], [4, "cmdoption-arg-gpus"], [4, "cmdoption-arg-group_size"], [4, "cmdoption-arg-hidden_size"], [4, "cmdoption-arg-htp_socs"], [4, "cmdoption-arg-ignored_scope"], [4, "cmdoption-arg-ignored_scope_type"], [4, "cmdoption-arg-input"], [4, "cmdoption-arg-input_dim"], [4, "cmdoption-arg-input_int32"], [4, "cmdoption-arg-input_layouts"], [4, "cmdoption-arg-input_model_dtype"], [4, "cmdoption-arg-input_name"], [4, "cmdoption-arg-input_names"], [4, "cmdoption-arg-input_nodes"], [4, "cmdoption-arg-input_shape"], [4, "cmdoption-arg-input_shapes"], [4, "cmdoption-arg-input_types"], [4, "cmdoption-arg-inputs_to_make_channel_last"], [4, "cmdoption-arg-inside_layer_modules"], [4, "cmdoption-arg-int4_accuracy_level"], [4, "cmdoption-arg-int4_block_size"], [4, "cmdoption-arg-inter_thread_num_list"], [4, "cmdoption-arg-intra_thread_num_list"], [4, "cmdoption-arg-io_bind"], [4, "cmdoption-arg-k_bits"], [4, "cmdoption-arg-k_clip_ratio"], [4, "cmdoption-arg-k_groupsize"], [4, "cmdoption-arg-keep_io_types"], [4, "cmdoption-arg-layer_name_filter"], [4, "cmdoption-arg-layers_block_name"], [4, "cmdoption-arg-lib_name"], [4, "cmdoption-arg-lib_targets"], [4, "cmdoption-arg-loftq_iter"], [4, "cmdoption-arg-logger"], [4, "cmdoption-arg-lora_alpha"], [4, "cmdoption-arg-lora_dropout"], [4, "cmdoption-arg-lora_r"], [4, "cmdoption-arg-make_inputs"], [4, "cmdoption-arg-max_finite_val"], [4, "cmdoption-arg-max_layer"], [4, "cmdoption-arg-merge_adapter_weights"], [4, "cmdoption-arg-metadata_only"], [4, "cmdoption-arg-metric"], [4, "cmdoption-arg-min_layer"], [4, "cmdoption-arg-min_positive_val"], [4, "cmdoption-arg-model_type"], [4, "cmdoption-arg-modules_to_fuse"], [4, "cmdoption-arg-modules_to_not_convert"], [4, "cmdoption-arg-modules_to_save"], [4, "cmdoption-arg-name_pattern"], [4, "cmdoption-arg-need_layer_fusing"], [4, "cmdoption-arg-no_repeat_ngram_size"], [4, "cmdoption-arg-node_block_list"], [4, "cmdoption-arg-nodes_to_exclude"], [4, "cmdoption-arg-nodes_to_quantize"], [4, "cmdoption-arg-nsamples"], [4, "cmdoption-arg-num_epochs"], [4, "cmdoption-arg-num_heads"], [4, "cmdoption-arg-num_key_value_heads"], [4, "cmdoption-arg-num_steps"], [4, "cmdoption-arg-only_onnxruntime"], [4, "cmdoption-arg-op_block_list"], [4, "cmdoption-arg-op_type_dict"], [4, "cmdoption-arg-op_types_to_quantize"], [4, "cmdoption-arg-opt_level"], [4, "cmdoption-arg-opt_level_list"], [4, "cmdoption-arg-optimization_options"], [4, "cmdoption-arg-optimize_model"], [4, "cmdoption-arg-optional_inputs"], [4, "cmdoption-arg-out_node"], [4, "cmdoption-arg-output_model"], [4, "cmdoption-arg-output_names"], [4, "cmdoption-arg-output_nodes"], [4, "cmdoption-arg-outputs_to_make_channel_last"], [4, "cmdoption-arg-outside_layer_modules"], [4, "cmdoption-arg-overrides_config"], [4, "cmdoption-arg-pack_inputs"], [4, "cmdoption-arg-pack_model_for_onnx_conversion"], [4, "cmdoption-arg-parallel_jobs"], [4, "cmdoption-arg-past_key_value_name"], [4, "cmdoption-arg-per_channel"], [4, "cmdoption-arg-percdamp"], [4, "cmdoption-arg-post"], [4, "cmdoption-arg-pre"], [4, "cmdoption-arg-precision"], [4, "cmdoption-arg-prepare_qnn_config"], [4, "cmdoption-arg-preset"], [4, "cmdoption-arg-providers_list"], [4, "cmdoption-arg-ptl_data_module"], [4, "cmdoption-arg-ptl_module"], [4, "cmdoption-arg-q_group_size"], [4, "cmdoption-arg-qconfig_func"], [4, "cmdoption-arg-qnn_extra_options"], [4, "cmdoption-arg-quant_format"], [4, "cmdoption-arg-quant_level"], [4, "cmdoption-arg-quant_mode"], [4, "cmdoption-arg-quant_preprocess"], [4, "cmdoption-arg-quant_type"], [4, "cmdoption-arg-recipes"], [4, "cmdoption-arg-reduce_range"], [4, "cmdoption-arg-rotate"], [4, "cmdoption-arg-rotation_seed"], [4, "cmdoption-arg-round_interval"], [4, "cmdoption-arg-save_as_external_data"], [4, "cmdoption-arg-save_format"], [4, "cmdoption-arg-save_metadata_for_token_generation"], [4, "cmdoption-arg-script_dir"], [4, "cmdoption-arg-search"], [4, "cmdoption-arg-seed"], [4, "cmdoption-arg-size_threshold"], [4, "cmdoption-arg-sparsity"], [4, "cmdoption-arg-static_groups"], [4, "cmdoption-arg-strict"], [4, "cmdoption-arg-sym"], [4, "cmdoption-arg-target_device"], [4, "cmdoption-arg-target_modules"], [4, "cmdoption-arg-target_opset"], [4, "cmdoption-arg-tool_command"], [4, "cmdoption-arg-tool_command_args"], [4, "cmdoption-arg-torch_dtype"], [4, "cmdoption-arg-train_data_config"], [4, "cmdoption-arg-training_args"], [4, "cmdoption-arg-training_loop_func"], [4, "cmdoption-arg-trt_fp16_enable"], [4, "cmdoption-arg-true_sequential"], [4, "cmdoption-arg-tuning_criterion"], [4, "cmdoption-arg-use_dynamo_exporter"], [4, "cmdoption-arg-use_enhanced_quantizer"], [4, "cmdoption-arg-use_external_data_format"], [4, "cmdoption-arg-use_forced_decoder_ids"], [4, "cmdoption-arg-use_gpu"], [4, "cmdoption-arg-use_gqa"], [4, "cmdoption-arg-use_logits_processor"], [4, "cmdoption-arg-use_prefix_vocab_mask"], [4, "cmdoption-arg-use_symbolic_shape_infer"], [4, "cmdoption-arg-use_temperature"], [4, "cmdoption-arg-use_vocab_mask"], [4, "cmdoption-arg-user_script"], [4, "cmdoption-arg-v_bits"], [4, "cmdoption-arg-v_clip_ratio"], [4, "cmdoption-arg-v_groupsize"], [4, "cmdoption-arg-val_data_config"], [4, "cmdoption-arg-version"], [4, "cmdoption-arg-w_asym"], [4, "cmdoption-arg-w_bit"], [4, "cmdoption-arg-w_bits"], [4, "cmdoption-arg-w_gptq"], [4, "cmdoption-arg-w_groupsize"], [4, "cmdoption-arg-w_rtn"], [4, "cmdoption-arg-weight_only_config"], [4, "cmdoption-arg-weight_type"], [4, "cmdoption-arg-workspace"], [4, "cmdoption-arg-zero_point"], [5, "cmdoption-arg-0"], [5, "cmdoption-arg-1"], [5, "cmdoption-arg-2"], [5, "cmdoption-arg-3"], [5, "cmdoption-arg-4"], [5, "cmdoption-arg-azureml_client"], [5, "cmdoption-arg-datastore_name"], [5, "cmdoption-arg-datastore_url"], [5, "cmdoption-arg-job_name"], [5, "cmdoption-arg-name"], [5, "cmdoption-arg-output_name"], [5, "cmdoption-arg-path"], [5, "cmdoption-arg-relative_path"], [5, "cmdoption-arg-version"], [6, "cmdoption-arg-0"], [6, "cmdoption-arg-1"], [6, "cmdoption-arg-group"], [6, "cmdoption-arg-multivariate"], [6, "cmdoption-arg-num_samples"], [6, "cmdoption-arg-seed"], [6, "cmdoption-arg-with_replacement"]], "components": [[4, "cmdoption-arg-components"]], "compress_to_fp16": [[4, "cmdoption-arg-compress_to_fp16"]], "compute_dtype": [[4, "cmdoption-arg-199"], [4, "cmdoption-arg-compute_dtype"]], "convert_attribute": [[4, "cmdoption-arg-10"], [4, "cmdoption-arg-104"], [4, "cmdoption-arg-111"], [4, "cmdoption-arg-131"], [4, "cmdoption-arg-15"], [4, "cmdoption-arg-152"], [4, "cmdoption-arg-171"], [4, "cmdoption-arg-177"], [4, "cmdoption-arg-183"], [4, "cmdoption-arg-188"], [4, "cmdoption-arg-227"], [4, "cmdoption-arg-24"], [4, "cmdoption-arg-262"], [4, "cmdoption-arg-29"], [4, "cmdoption-arg-35"], [4, "cmdoption-arg-40"], [4, "cmdoption-arg-45"], [4, "cmdoption-arg-5"], [4, "cmdoption-arg-50"], [4, "cmdoption-arg-72"], [4, "cmdoption-arg-99"], [4, "cmdoption-arg-convert_attribute"]], "cpu_cores": [[4, "cmdoption-arg-cpu_cores"]], "create_training_args() (olive.passes.pytorch.lora.hftrainingarguments method)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.create_training_args"]], "damp_percent": [[4, "cmdoption-arg-damp_percent"]], "data_config": [[4, "cmdoption-arg-125"], [4, "cmdoption-arg-145"], [4, "cmdoption-arg-154"], [4, "cmdoption-arg-216"], [4, "cmdoption-arg-219"], [4, "cmdoption-arg-232"], [4, "cmdoption-arg-242"], [4, "cmdoption-arg-246"], [4, "cmdoption-arg-251"], [4, "cmdoption-arg-60"], [4, "cmdoption-arg-82"], [4, "cmdoption-arg-data_config"]], "desc_act": [[4, "cmdoption-arg-desc_act"]], "device": [[4, "cmdoption-arg-105"], [4, "cmdoption-arg-113"], [4, "cmdoption-arg-133"], [4, "cmdoption-arg-18"], [4, "cmdoption-arg-231"], [4, "cmdoption-arg-256"], [4, "cmdoption-arg-device"]], "dim_param": [[4, "cmdoption-arg-dim_param"]], "dim_value": [[4, "cmdoption-arg-dim_value"]], "do_validate": [[4, "cmdoption-arg-do_validate"]], "domain": [[4, "cmdoption-arg-115"], [4, "cmdoption-arg-135"], [4, "cmdoption-arg-domain"]], "double_quant": [[4, "cmdoption-arg-double_quant"]], "duo_scaling": [[4, "cmdoption-arg-duo_scaling"]], "dynamic_lora_r": [[4, "cmdoption-arg-dynamic_lora_r"]], "element_wise_binary_ops": [[4, "cmdoption-arg-element_wise_binary_ops"]], "enable_cuda_graph": [[4, "cmdoption-arg-263"], [4, "cmdoption-arg-enable_cuda_graph"]], "enable_dpu": [[4, "cmdoption-arg-enable_dpu"]], "enable_htp": [[4, "cmdoption-arg-enable_htp"]], "enable_profiling": [[4, "cmdoption-arg-enable_profiling"]], "eval_data_config": [[4, "cmdoption-arg-197"], [4, "cmdoption-arg-207"], [4, "cmdoption-arg-eval_data_config"]], "evaluation_strategy (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.evaluation_strategy"]], "example_input_func": [[4, "cmdoption-arg-example_input_func"]], "exclude_embeds": [[4, "cmdoption-arg-exclude_embeds"]], "exclude_lm_head": [[4, "cmdoption-arg-exclude_lm_head"]], "excluded_precisions": [[4, "cmdoption-arg-120"], [4, "cmdoption-arg-140"], [4, "cmdoption-arg-excluded_precisions"]], "execution_mode_list": [[4, "cmdoption-arg-execution_mode_list"]], "export_compatible": [[4, "cmdoption-arg-export_compatible"]], "external_data_name": [[4, "cmdoption-arg-102"], [4, "cmdoption-arg-109"], [4, "cmdoption-arg-129"], [4, "cmdoption-arg-13"], [4, "cmdoption-arg-150"], [4, "cmdoption-arg-169"], [4, "cmdoption-arg-175"], [4, "cmdoption-arg-181"], [4, "cmdoption-arg-186"], [4, "cmdoption-arg-22"], [4, "cmdoption-arg-225"], [4, "cmdoption-arg-260"], [4, "cmdoption-arg-27"], [4, "cmdoption-arg-3"], [4, "cmdoption-arg-33"], [4, "cmdoption-arg-38"], [4, "cmdoption-arg-43"], [4, "cmdoption-arg-48"], [4, "cmdoption-arg-70"], [4, "cmdoption-arg-8"], [4, "cmdoption-arg-97"], [4, "cmdoption-arg-external_data_name"]], "extra.sigmoid.nnapi": [[4, "cmdoption-arg-61"], [4, "cmdoption-arg-88"], [4, "cmdoption-arg-extra.Sigmoid.nnapi"]], "extra_args": [[4, "cmdoption-arg-220"], [4, "cmdoption-arg-228"], [4, "cmdoption-arg-230"], [4, "cmdoption-arg-257"], [4, "cmdoption-arg-extra_args"]], "extra_args (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.extra_args"]], "extra_configs": [[4, "cmdoption-arg-218"], [4, "cmdoption-arg-extra_configs"]], "extra_options": [[4, "cmdoption-arg-166"], [4, "cmdoption-arg-67"], [4, "cmdoption-arg-94"], [4, "cmdoption-arg-extra_options"]], "extra_session_config": [[4, "cmdoption-arg-extra_session_config"]], "final_orientation": [[4, "cmdoption-arg-final_orientation"]], "float16": [[4, "cmdoption-arg-250"], [4, "cmdoption-arg-float16"]], "force_evaluate_other_eps": [[4, "cmdoption-arg-force_evaluate_other_eps"]], "force_fp16_inputs": [[4, "cmdoption-arg-force_fp16_inputs"]], "force_fp32_nodes": [[4, "cmdoption-arg-force_fp32_nodes"]], "force_fp32_ops": [[4, "cmdoption-arg-force_fp32_ops"]], "fp16": [[4, "cmdoption-arg-255"], [4, "cmdoption-arg-fp16"]], "fuse_layernorm": [[4, "cmdoption-arg-fuse_layernorm"]], "gptq_damping": [[4, "cmdoption-arg-gptq_damping"]], "gptq_opt_scales": [[4, "cmdoption-arg-gptq_opt_scales"]], "gpus": [[4, "cmdoption-arg-gpus"]], "gradient_checkpointing (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.gradient_checkpointing"]], "group_size": [[4, "cmdoption-arg-group_size"]], "hidden_size": [[4, "cmdoption-arg-hidden_size"]], "htp_socs": [[4, "cmdoption-arg-htp_socs"]], "ignored_scope": [[4, "cmdoption-arg-ignored_scope"]], "ignored_scope_type": [[4, "cmdoption-arg-ignored_scope_type"]], "input": [[4, "cmdoption-arg-input"]], "input_dim": [[4, "cmdoption-arg-input_dim"]], "input_int32": [[4, "cmdoption-arg-input_int32"]], "input_layouts": [[4, "cmdoption-arg-input_layouts"]], "input_model_dtype": [[4, "cmdoption-arg-245"], [4, "cmdoption-arg-input_model_dtype"]], "input_name": [[4, "cmdoption-arg-input_name"]], "input_names": [[4, "cmdoption-arg-input_names"]], "input_nodes": [[4, "cmdoption-arg-input_nodes"]], "input_shape": [[4, "cmdoption-arg-input_shape"]], "input_shapes": [[4, "cmdoption-arg-input_shapes"]], "input_types": [[4, "cmdoption-arg-input_types"]], "inputs_to_make_channel_last": [[4, "cmdoption-arg-inputs_to_make_channel_last"]], "inside_layer_modules": [[4, "cmdoption-arg-inside_layer_modules"]], "int4_accuracy_level": [[4, "cmdoption-arg-int4_accuracy_level"]], "int4_block_size": [[4, "cmdoption-arg-int4_block_size"]], "inter_thread_num_list": [[4, "cmdoption-arg-inter_thread_num_list"]], "intra_thread_num_list": [[4, "cmdoption-arg-intra_thread_num_list"]], "io_bind": [[4, "cmdoption-arg-io_bind"]], "k_bits": [[4, "cmdoption-arg-k_bits"]], "k_clip_ratio": [[4, "cmdoption-arg-k_clip_ratio"]], "k_groupsize": [[4, "cmdoption-arg-k_groupsize"]], "keep_io_types": [[4, "cmdoption-arg-19"], [4, "cmdoption-arg-keep_io_types"]], "layer_name_filter": [[4, "cmdoption-arg-249"], [4, "cmdoption-arg-layer_name_filter"]], "layers_block_name": [[4, "cmdoption-arg-layers_block_name"]], "learning_rate (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.learning_rate"]], "lib_name": [[4, "cmdoption-arg-lib_name"]], "lib_targets": [[4, "cmdoption-arg-lib_targets"]], "loftq_iter": [[4, "cmdoption-arg-loftq_iter"]], "logger": [[4, "cmdoption-arg-logger"]], "lora_alpha": [[4, "cmdoption-arg-191"], [4, "cmdoption-arg-201"], [4, "cmdoption-arg-lora_alpha"]], "lora_dropout": [[4, "cmdoption-arg-192"], [4, "cmdoption-arg-202"], [4, "cmdoption-arg-lora_dropout"]], "lora_r": [[4, "cmdoption-arg-190"], [4, "cmdoption-arg-200"], [4, "cmdoption-arg-lora_r"]], "lr_scheduler_type (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.lr_scheduler_type"]], "make_inputs": [[4, "cmdoption-arg-make_inputs"]], "max_finite_val": [[4, "cmdoption-arg-max_finite_val"]], "max_layer": [[4, "cmdoption-arg-248"], [4, "cmdoption-arg-max_layer"]], "merge_adapter_weights": [[4, "cmdoption-arg-merge_adapter_weights"]], "metadata_only": [[4, "cmdoption-arg-metadata_only"]], "metric": [[4, "cmdoption-arg-122"], [4, "cmdoption-arg-142"], [4, "cmdoption-arg-metric"]], "min_layer": [[4, "cmdoption-arg-247"], [4, "cmdoption-arg-min_layer"]], "min_positive_val": [[4, "cmdoption-arg-min_positive_val"]], "model_type": [[4, "cmdoption-arg-217"], [4, "cmdoption-arg-model_type"]], "modules_to_fuse": [[4, "cmdoption-arg-modules_to_fuse"]], "modules_to_not_convert": [[4, "cmdoption-arg-modules_to_not_convert"]], "modules_to_save": [[4, "cmdoption-arg-193"], [4, "cmdoption-arg-203"], [4, "cmdoption-arg-modules_to_save"]], "name_pattern": [[4, "cmdoption-arg-name_pattern"]], "need_layer_fusing": [[4, "cmdoption-arg-need_layer_fusing"]], "no_repeat_ngram_size": [[4, "cmdoption-arg-no_repeat_ngram_size"]], "node_block_list": [[4, "cmdoption-arg-node_block_list"]], "nodes_to_exclude": [[4, "cmdoption-arg-158"], [4, "cmdoption-arg-56"], [4, "cmdoption-arg-78"], [4, "cmdoption-arg-nodes_to_exclude"]], "nodes_to_quantize": [[4, "cmdoption-arg-157"], [4, "cmdoption-arg-55"], [4, "cmdoption-arg-77"], [4, "cmdoption-arg-nodes_to_quantize"]], "nsamples": [[4, "cmdoption-arg-nsamples"]], "num_epochs": [[4, "cmdoption-arg-num_epochs"]], "num_heads": [[4, "cmdoption-arg-num_heads"]], "num_key_value_heads": [[4, "cmdoption-arg-num_key_value_heads"]], "num_steps": [[4, "cmdoption-arg-num_steps"]], "only_onnxruntime": [[4, "cmdoption-arg-only_onnxruntime"]], "op_block_list": [[4, "cmdoption-arg-30"], [4, "cmdoption-arg-op_block_list"]], "op_type_dict": [[4, "cmdoption-arg-124"], [4, "cmdoption-arg-144"], [4, "cmdoption-arg-op_type_dict"]], "op_types_to_quantize": [[4, "cmdoption-arg-156"], [4, "cmdoption-arg-53"], [4, "cmdoption-arg-75"], [4, "cmdoption-arg-op_types_to_quantize"]], "opt_level": [[4, "cmdoption-arg-opt_level"]], "opt_level_list": [[4, "cmdoption-arg-opt_level_list"]], "optim (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.optim"]], "optimization_options": [[4, "cmdoption-arg-optimization_options"]], "optimize_model": [[4, "cmdoption-arg-optimize_model"]], "optional_inputs": [[4, "cmdoption-arg-optional_inputs"]], "out_node": [[4, "cmdoption-arg-out_node"]], "output_dir (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.output_dir"]], "output_model": [[4, "cmdoption-arg-output_model"]], "output_names": [[4, "cmdoption-arg-output_names"]], "output_nodes": [[4, "cmdoption-arg-output_nodes"]], "outputs_to_make_channel_last": [[4, "cmdoption-arg-outputs_to_make_channel_last"]], "outside_layer_modules": [[4, "cmdoption-arg-outside_layer_modules"]], "overrides_config": [[4, "cmdoption-arg-overrides_config"]], "overwrite_output_dir (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.overwrite_output_dir"]], "pack_inputs": [[4, "cmdoption-arg-pack_inputs"]], "pack_model_for_onnx_conversion": [[4, "cmdoption-arg-pack_model_for_onnx_conversion"]], "parallel_jobs": [[4, "cmdoption-arg-parallel_jobs"]], "past_key_value_name": [[4, "cmdoption-arg-past_key_value_name"]], "per_channel": [[4, "cmdoption-arg-159"], [4, "cmdoption-arg-57"], [4, "cmdoption-arg-79"], [4, "cmdoption-arg-per_channel"]], "percdamp": [[4, "cmdoption-arg-percdamp"]], "post": [[4, "cmdoption-arg-post"]], "pre": [[4, "cmdoption-arg-pre"]], "precision": [[4, "cmdoption-arg-precision"]], "prepare_qnn_config": [[4, "cmdoption-arg-86"], [4, "cmdoption-arg-prepare_qnn_config"]], "preset": [[4, "cmdoption-arg-preset"]], "providers_list": [[4, "cmdoption-arg-providers_list"]], "ptl_data_module": [[4, "cmdoption-arg-ptl_data_module"]], "ptl_module": [[4, "cmdoption-arg-ptl_module"]], "q_group_size": [[4, "cmdoption-arg-q_group_size"]], "qconfig_func": [[4, "cmdoption-arg-qconfig_func"]], "qnn_extra_options": [[4, "cmdoption-arg-87"], [4, "cmdoption-arg-qnn_extra_options"]], "quant_format": [[4, "cmdoption-arg-126"], [4, "cmdoption-arg-146"], [4, "cmdoption-arg-162"], [4, "cmdoption-arg-84"], [4, "cmdoption-arg-quant_format"]], "quant_level": [[4, "cmdoption-arg-119"], [4, "cmdoption-arg-139"], [4, "cmdoption-arg-quant_level"]], "quant_mode": [[4, "cmdoption-arg-153"], [4, "cmdoption-arg-51"], [4, "cmdoption-arg-73"], [4, "cmdoption-arg-quant_mode"]], "quant_preprocess": [[4, "cmdoption-arg-160"], [4, "cmdoption-arg-59"], [4, "cmdoption-arg-81"], [4, "cmdoption-arg-quant_preprocess"]], "quant_type": [[4, "cmdoption-arg-quant_type"]], "recipes": [[4, "cmdoption-arg-117"], [4, "cmdoption-arg-137"], [4, "cmdoption-arg-recipes"]], "reduce_range": [[4, "cmdoption-arg-106"], [4, "cmdoption-arg-118"], [4, "cmdoption-arg-138"], [4, "cmdoption-arg-58"], [4, "cmdoption-arg-80"], [4, "cmdoption-arg-reduce_range"]], "report_to (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.report_to"]], "resume_from_checkpoint (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.resume_from_checkpoint"]], "rotate": [[4, "cmdoption-arg-rotate"]], "rotation_seed": [[4, "cmdoption-arg-rotation_seed"]], "round_interval": [[4, "cmdoption-arg-round_interval"]], "save_as_external_data": [[4, "cmdoption-arg-1"], [4, "cmdoption-arg-100"], [4, "cmdoption-arg-107"], [4, "cmdoption-arg-11"], [4, "cmdoption-arg-127"], [4, "cmdoption-arg-148"], [4, "cmdoption-arg-167"], [4, "cmdoption-arg-173"], [4, "cmdoption-arg-179"], [4, "cmdoption-arg-184"], [4, "cmdoption-arg-20"], [4, "cmdoption-arg-223"], [4, "cmdoption-arg-25"], [4, "cmdoption-arg-258"], [4, "cmdoption-arg-31"], [4, "cmdoption-arg-36"], [4, "cmdoption-arg-41"], [4, "cmdoption-arg-46"], [4, "cmdoption-arg-6"], [4, "cmdoption-arg-68"], [4, "cmdoption-arg-95"], [4, "cmdoption-arg-save_as_external_data"]], "save_format": [[4, "cmdoption-arg-save_format"]], "save_metadata_for_token_generation": [[4, "cmdoption-arg-save_metadata_for_token_generation"]], "script_dir": [[4, "cmdoption-arg-17"], [4, "cmdoption-arg-210"], [4, "cmdoption-arg-213"], [4, "cmdoption-arg-215"], [4, "cmdoption-arg-240"], [4, "cmdoption-arg-244"], [4, "cmdoption-arg-253"], [4, "cmdoption-arg-script_dir"]], "search": [[4, "cmdoption-arg-search"]], "seed": [[4, "cmdoption-arg-233"], [4, "cmdoption-arg-238"], [4, "cmdoption-arg-241"], [4, "cmdoption-arg-seed"], [6, "cmdoption-arg-1"], [6, "cmdoption-arg-seed"]], "size_threshold": [[4, "cmdoption-arg-103"], [4, "cmdoption-arg-110"], [4, "cmdoption-arg-130"], [4, "cmdoption-arg-14"], [4, "cmdoption-arg-151"], [4, "cmdoption-arg-170"], [4, "cmdoption-arg-176"], [4, "cmdoption-arg-182"], [4, "cmdoption-arg-187"], [4, "cmdoption-arg-226"], [4, "cmdoption-arg-23"], [4, "cmdoption-arg-261"], [4, "cmdoption-arg-28"], [4, "cmdoption-arg-34"], [4, "cmdoption-arg-39"], [4, "cmdoption-arg-4"], [4, "cmdoption-arg-44"], [4, "cmdoption-arg-49"], [4, "cmdoption-arg-71"], [4, "cmdoption-arg-9"], [4, "cmdoption-arg-98"], [4, "cmdoption-arg-size_threshold"]], "sparsity": [[4, "cmdoption-arg-234"], [4, "cmdoption-arg-sparsity"]], "static_groups": [[4, "cmdoption-arg-static_groups"]], "strict": [[4, "cmdoption-arg-strict"]], "sym": [[4, "cmdoption-arg-sym"]], "target_device": [[4, "cmdoption-arg-221"], [4, "cmdoption-arg-target_device"]], "target_modules": [[4, "cmdoption-arg-target_modules"]], "target_opset": [[4, "cmdoption-arg-0"], [4, "cmdoption-arg-172"], [4, "cmdoption-arg-222"], [4, "cmdoption-arg-254"], [4, "cmdoption-arg-target_opset"]], "tool_command": [[4, "cmdoption-arg-tool_command"]], "tool_command_args": [[4, "cmdoption-arg-tool_command_args"]], "torch_dtype": [[4, "cmdoption-arg-189"], [4, "cmdoption-arg-194"], [4, "cmdoption-arg-204"], [4, "cmdoption-arg-torch_dtype"]], "train_data_config": [[4, "cmdoption-arg-196"], [4, "cmdoption-arg-206"], [4, "cmdoption-arg-211"], [4, "cmdoption-arg-train_data_config"]], "training_args": [[4, "cmdoption-arg-198"], [4, "cmdoption-arg-208"], [4, "cmdoption-arg-training_args"]], "training_loop_func": [[4, "cmdoption-arg-training_loop_func"]], "trt_fp16_enable": [[4, "cmdoption-arg-trt_fp16_enable"]], "true_sequential": [[4, "cmdoption-arg-true_sequential"]], "tuning_criterion": [[4, "cmdoption-arg-121"], [4, "cmdoption-arg-141"], [4, "cmdoption-arg-tuning_criterion"]], "use_dynamo_exporter": [[4, "cmdoption-arg-use_dynamo_exporter"]], "use_enhanced_quantizer": [[4, "cmdoption-arg-use_enhanced_quantizer"]], "use_external_data_format": [[4, "cmdoption-arg-use_external_data_format"]], "use_forced_decoder_ids": [[4, "cmdoption-arg-use_forced_decoder_ids"]], "use_gpu": [[4, "cmdoption-arg-178"], [4, "cmdoption-arg-use_gpu"]], "use_gqa": [[4, "cmdoption-arg-use_gqa"]], "use_logits_processor": [[4, "cmdoption-arg-use_logits_processor"]], "use_prefix_vocab_mask": [[4, "cmdoption-arg-use_prefix_vocab_mask"]], "use_symbolic_shape_infer": [[4, "cmdoption-arg-use_symbolic_shape_infer"]], "use_temperature": [[4, "cmdoption-arg-use_temperature"]], "use_vocab_mask": [[4, "cmdoption-arg-use_vocab_mask"]], "user_script": [[4, "cmdoption-arg-16"], [4, "cmdoption-arg-209"], [4, "cmdoption-arg-212"], [4, "cmdoption-arg-214"], [4, "cmdoption-arg-239"], [4, "cmdoption-arg-243"], [4, "cmdoption-arg-252"], [4, "cmdoption-arg-user_script"]], "v_bits": [[4, "cmdoption-arg-v_bits"]], "v_clip_ratio": [[4, "cmdoption-arg-v_clip_ratio"]], "v_groupsize": [[4, "cmdoption-arg-v_groupsize"]], "val_data_config": [[4, "cmdoption-arg-val_data_config"]], "version": [[4, "cmdoption-arg-version"], [5, "cmdoption-arg-version"]], "w_asym": [[4, "cmdoption-arg-w_asym"]], "w_bit": [[4, "cmdoption-arg-w_bit"]], "w_bits": [[4, "cmdoption-arg-w_bits"]], "w_gptq": [[4, "cmdoption-arg-w_gptq"]], "w_groupsize": [[4, "cmdoption-arg-w_groupsize"]], "w_rtn": [[4, "cmdoption-arg-w_rtn"]], "warmup_ratio (olive.passes.pytorch.lora.hftrainingarguments attribute)": [[4, "olive.passes.pytorch.lora.HFTrainingArguments.warmup_ratio"]], "weight_only_config": [[4, "cmdoption-arg-123"], [4, "cmdoption-arg-143"], [4, "cmdoption-arg-weight_only_config"]], "weight_type": [[4, "cmdoption-arg-155"], [4, "cmdoption-arg-52"], [4, "cmdoption-arg-74"], [4, "cmdoption-arg-weight_type"]], "workspace": [[4, "cmdoption-arg-116"], [4, "cmdoption-arg-136"], [4, "cmdoption-arg-workspace"]], "zero_point": [[4, "cmdoption-arg-zero_point"]], "azuremldatastore (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.AzureMLDatastore"]], "azuremljoboutput (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.AzureMLJobOutput"]], "azuremlmodel (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.AzureMLModel"]], "azuremlregistrymodel (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.AzureMLRegistryModel"]], "localfile (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.LocalFile"]], "localfolder (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.LocalFolder"]], "resourcetype (class in olive.resource_path)": [[5, "olive.resource_path.ResourceType"]], "stringname (olive.resource_path.resourcetype attribute)": [[5, "olive.resource_path.ResourceType.StringName"]], "aml_config_path (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.aml_config_path"]], "azureml_client": [[5, "cmdoption-arg-2"], [5, "cmdoption-arg-3"], [5, "cmdoption-arg-azureml_client"]], "create_client() (olive.azureml.azureml_client.azuremlclientconfig method)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.create_client"]], "create_registry_client() (olive.azureml.azureml_client.azuremlclientconfig method)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.create_registry_client"]], "datastore_name": [[5, "cmdoption-arg-datastore_name"]], "datastore_url": [[5, "cmdoption-arg-datastore_url"]], "default_auth_params (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.default_auth_params"]], "get_workspace_config() (olive.azureml.azureml_client.azuremlclientconfig method)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.get_workspace_config"]], "job_name": [[5, "cmdoption-arg-job_name"]], "keyvault_name (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.keyvault_name"]], "max_operation_retries (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.max_operation_retries"]], "name": [[5, "cmdoption-arg-1"], [5, "cmdoption-arg-name"]], "operation_retry_interval (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.operation_retry_interval"]], "output_name": [[5, "cmdoption-arg-output_name"]], "path": [[5, "cmdoption-arg-0"], [5, "cmdoption-arg-path"]], "read_timeout (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.read_timeout"]], "relative_path": [[5, "cmdoption-arg-4"], [5, "cmdoption-arg-relative_path"]], "resource_group (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.resource_group"]], "subscription_id (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.subscription_id"]], "workspace_name (olive.azureml.azureml_client.azuremlclientconfig attribute)": [[5, "olive.azureml.azureml_client.AzureMLClientConfig.workspace_name"]], "group": [[6, "cmdoption-arg-group"]], "multivariate": [[6, "cmdoption-arg-multivariate"]], "num_samples": [[6, "cmdoption-arg-0"], [6, "cmdoption-arg-num_samples"]], "with_replacement": [[6, "cmdoption-arg-with_replacement"]], "acceleratorconfig (class in olive.systems.common)": [[7, "olive.systems.common.AcceleratorConfig"]], "azureml (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.AzureML"]], "azuremlsystem (class in olive.systems.azureml)": [[7, "olive.systems.azureml.AzureMLSystem"]], "azurend12ssystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND12SSystem"]], "azurend24rssystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND24RSSystem"]], "azurend24ssystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND24SSystem"]], "azurend6ssystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND6SSystem"]], "azurend96a100system (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND96A100System"]], "azurend96asystem (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureND96ASystem"]], "azurendv2system (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.AzureNDV2System"]], "cpu (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.CPU"]], "cpu_spr (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.CPU_SPR"]], "device (class in olive.hardware.accelerator)": [[7, "olive.hardware.accelerator.Device"]], "docker (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.Docker"]], "dockersystem (class in olive.systems.docker)": [[7, "olive.systems.docker.DockerSystem"]], "gpu (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.GPU"]], "intel_myriad (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.INTEL_MYRIAD"]], "isolatedort (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.IsolatedORT"]], "isolatedortsystem (class in olive.systems.isolated_ort)": [[7, "olive.systems.isolated_ort.IsolatedORTSystem"]], "local (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.Local"]], "localsystem (class in olive.systems.local)": [[7, "olive.systems.local.LocalSystem"]], "npu (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.NPU"]], "pythonenvironment (olive.systems.common.systemtype attribute)": [[7, "olive.systems.common.SystemType.PythonEnvironment"]], "pythonenvironmentsystem (class in olive.systems.python_environment)": [[7, "olive.systems.python_environment.PythonEnvironmentSystem"]], "surfaceprosystem1796 (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.SurfaceProSystem1796"]], "surfaceprosystem1807 (class in olive.systems.system_alias)": [[7, "olive.systems.system_alias.SurfaceProSystem1807"]], "systemtype (class in olive.systems.common)": [[7, "olive.systems.common.SystemType"]], "vpu (olive.hardware.accelerator.device attribute)": [[7, "olive.hardware.accelerator.Device.VPU"]], "accelerators (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.accelerators"]], "accelerators (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.accelerators"]], "accelerators (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.accelerators"]], "accelerators (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.accelerators"]], "accelerators (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.accelerators"]], "accelerators (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.accelerators"]], "accelerators (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.accelerators"]], "accelerators (olive.systems.system_alias.surfaceprosystem1796 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1796.accelerators"]], "accelerators (olive.systems.system_alias.surfaceprosystem1807 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1807.accelerators"]], "accelerators (olive.systems.system_config.localtargetuserconfig attribute)": [[7, "olive.systems.system_config.LocalTargetUserConfig.accelerators"]], "aml_compute (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.aml_compute"]], "aml_docker_config (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.aml_docker_config"]], "aml_environment_config (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.aml_environment_config"]], "azureml_client_config (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.azureml_client_config"]], "base_image (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.base_image"]], "build_args (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.build_args"]], "build_context_path (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.build_context_path"]], "build_context_path (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.build_context_path"]], "conda_file_path (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.conda_file_path"]], "config (olive.systems.system_config.systemconfig attribute)": [[7, "olive.systems.system_config.SystemConfig.config"]], "create_system() (olive.systems.system_config.systemconfig method)": [[7, "olive.systems.system_config.SystemConfig.create_system"]], "datastores (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.datastores"]], "device (olive.systems.common.acceleratorconfig attribute)": [[7, "olive.systems.common.AcceleratorConfig.device"]], "dockerfile (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.dockerfile"]], "dockerfile (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.dockerfile"]], "environment_variables (olive.systems.system_config.isolatedorttargetuserconfig attribute)": [[7, "olive.systems.system_config.IsolatedORTTargetUserConfig.environment_variables"]], "execution_providers (olive.systems.common.acceleratorconfig attribute)": [[7, "olive.systems.common.AcceleratorConfig.execution_providers"]], "hf_token (olive.systems.system_config.localtargetuserconfig attribute)": [[7, "olive.systems.system_config.LocalTargetUserConfig.hf_token"]], "image_name (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.image_name"]], "instance_count (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.instance_count"]], "is_dev (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.is_dev"]], "is_dev (olive.systems.system_config.dockertargetuserconfig attribute)": [[7, "olive.systems.system_config.DockerTargetUserConfig.is_dev"]], "local_docker_config (olive.systems.system_config.dockertargetuserconfig attribute)": [[7, "olive.systems.system_config.DockerTargetUserConfig.local_docker_config"]], "module": [[7, "module-olive.systems.system_alias"]], "name (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.name"]], "num_cpus (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.num_cpus"]], "num_cpus (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.num_cpus"]], "num_cpus (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.num_cpus"]], "num_gpus (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.num_gpus"]], "num_gpus (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.num_gpus"]], "num_gpus (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.num_gpus"]], "olive.systems.system_alias": [[7, "module-olive.systems.system_alias"]], "olive_managed_env (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.olive_managed_env"]], "olive_managed_env (olive.systems.system_config.dockertargetuserconfig attribute)": [[7, "olive.systems.system_config.DockerTargetUserConfig.olive_managed_env"]], "olive_managed_env (olive.systems.system_config.pythonenvironmenttargetuserconfig attribute)": [[7, "olive.systems.system_config.PythonEnvironmentTargetUserConfig.olive_managed_env"]], "olive_managed_env (olive.systems.system_config.systemconfig property)": [[7, "olive.systems.system_config.SystemConfig.olive_managed_env"]], "prepend_to_path (olive.systems.system_config.isolatedorttargetuserconfig attribute)": [[7, "olive.systems.system_config.IsolatedORTTargetUserConfig.prepend_to_path"]], "python_environment_path (olive.systems.system_config.isolatedorttargetuserconfig attribute)": [[7, "olive.systems.system_config.IsolatedORTTargetUserConfig.python_environment_path"]], "requirements_file (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.requirements_file"]], "requirements_file (olive.systems.system_config.dockertargetuserconfig attribute)": [[7, "olive.systems.system_config.DockerTargetUserConfig.requirements_file"]], "requirements_file (olive.systems.system_config.pythonenvironmenttargetuserconfig attribute)": [[7, "olive.systems.system_config.PythonEnvironmentTargetUserConfig.requirements_file"]], "resources (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.resources"]], "run_params (olive.systems.docker.localdockerconfig attribute)": [[7, "olive.systems.docker.LocalDockerConfig.run_params"]], "sku (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.sku"]], "sku (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.sku"]], "sku (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.sku"]], "sku (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.sku"]], "sku (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.sku"]], "sku (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.sku"]], "sku (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.sku"]], "sku (olive.systems.system_alias.surfaceprosystem1796 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1796.sku"]], "sku (olive.systems.system_alias.surfaceprosystem1807 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1807.sku"]], "system_type (olive.systems.system_alias.azurend12ssystem attribute)": [[7, "olive.systems.system_alias.AzureND12SSystem.system_type"]], "system_type (olive.systems.system_alias.azurend24rssystem attribute)": [[7, "olive.systems.system_alias.AzureND24RSSystem.system_type"]], "system_type (olive.systems.system_alias.azurend24ssystem attribute)": [[7, "olive.systems.system_alias.AzureND24SSystem.system_type"]], "system_type (olive.systems.system_alias.azurend6ssystem attribute)": [[7, "olive.systems.system_alias.AzureND6SSystem.system_type"]], "system_type (olive.systems.system_alias.azurend96a100system attribute)": [[7, "olive.systems.system_alias.AzureND96A100System.system_type"]], "system_type (olive.systems.system_alias.azurend96asystem attribute)": [[7, "olive.systems.system_alias.AzureND96ASystem.system_type"]], "system_type (olive.systems.system_alias.azurendv2system attribute)": [[7, "olive.systems.system_alias.AzureNDV2System.system_type"]], "system_type (olive.systems.system_alias.surfaceprosystem1796 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1796.system_type"]], "system_type (olive.systems.system_alias.surfaceprosystem1807 attribute)": [[7, "olive.systems.system_alias.SurfaceProSystem1807.system_type"]], "tags (olive.systems.system_config.azuremltargetuserconfig attribute)": [[7, "olive.systems.system_config.AzureMLTargetUserConfig.tags"]], "type (olive.systems.system_config.systemconfig attribute)": [[7, "olive.systems.system_config.SystemConfig.type"]], "version (olive.systems.azureml.azuremldockerconfig attribute)": [[7, "olive.systems.azureml.AzureMLDockerConfig.version"]]}})
\ No newline at end of file