Buckets:

hf-doc-build
/

doc-dev

Files

xet

hf-doc-build/doc-dev / transformers /pr_33892 /en /quantization /compressed_tensors.html

rtrm

about 1 month ago

download

raw

26.7 kB

	<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"compressed-tensors","local":"compressed-tensors","sections":[{"title":"Model checkpoint","local":"model-checkpoint","sections":[],"depth":2}],"depth":1}">
	<link href="/docs/transformers/pr_33892/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/entry/start.b2c4257a.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/scheduler.31fdf58d.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/singletons.9860629f.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/index.252883d5.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/paths.e85c0ec8.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/entry/app.05ef1f97.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/preload-helper.40847a0e.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/index.2f76fdf0.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/nodes/0.ca4aafa4.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/each.e59479a4.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/nodes/526.5f0c09fd.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/CopyLLMTxtMenu.ff482081.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/MermaidChart.svelte_svelte_type_style_lang.71f274cc.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/IconCopy.ac192424.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/CodeBlock.ab12f8e1.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33892/en/_app/immutable/chunks/HfOption.fb051768.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{"title":"compressed-tensors","local":"compressed-tensors","sections":[{"title":"Model checkpoint","local":"model-checkpoint","sections":[],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <div class="items-center shrink-0 min-w-[100px] max-sm:min-w-[50px] justify-end ml-auto flex" style="float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"><div class="inline-flex rounded-md max-sm:rounded-sm"><button class="inline-flex items-center gap-1 max-sm:gap-0.5 h-6 max-sm:h-5 px-2 max-sm:px-1.5 text-[11px] max-sm:text-[9px] font-medium text-gray-800 border border-r-0 rounded-l-md max-sm:rounded-l-sm border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-live="polite"><span class="inline-flex items-center justify-center rounded-md p-0.5 max-sm:p-0"><svg class="w-3 h-3 max-sm:w-2.5 max-sm:h-2.5" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg></span> <span>Copy page</span></button> <button class="inline-flex items-center justify-center w-6 max-sm:w-5 h-6 max-sm:h-5 disabled:pointer-events-none text-sm text-gray-500 hover:text-gray-700 dark:hover:text-white rounded-r-md max-sm:rounded-r-sm border border-l transition border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-haspopup="menu" aria-expanded="false" aria-label="Open copy menu"><svg class="transition-transform text-gray-400 overflow-visible w-3 h-3 max-sm:w-2.5 max-sm:h-2.5 rotate-0" width="1em" height="1em" viewBox="0 0 12 7" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M1 1L6 6L11 1" stroke="currentColor"></path></svg></button></div> </div> <h1 class="relative group"><a id="compressed-tensors" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#compressed-tensors"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>compressed-tensors</span></h1> <p data-svelte-h="svelte-1625r46"><a href="https://github.com/neuralmagic/compressed-tensors" rel="nofollow">compressed-tensors</a> extends <a href="https://github.com/huggingface/safetensors" rel="nofollow">safetensors</a> files to compressed tensor data types to provide a unified checkpoint format for storing and loading various quantization and sparsity formats such dense, int-quantized (int8), float-quantized (fp8), and pack-quantized (int4 or int8 weight-quantized packed into int32).</p> <p data-svelte-h="svelte-haujfi">compressed-tensors supports fine-tuning with <a href="https://huggingface.co/docs/peft" rel="nofollow">PEFT</a> and includes the following features as well.</p> <ul data-svelte-h="svelte-1usmb9k"><li>fp8, int4, int8 weight and activation precisions.</li> <li>Quantization scales and zero-points strategies for <a href="https://github.com/neuralmagic/compressed-tensors/blob/83b2e7a969d70606421a76b9a3d112646077c8de/src/compressed_tensors/quantization/quant_args.py#L43-L52" rel="nofollow">tensor, channel, group, block, token</a>.</li> <li>Dynamic per-token activation quantization (or any static strategy).</li> <li>Weight sparsity (unstructured or semi-structured like 2:4) can be composed with quantization for extreme compression.</li> <li>Quantization of arbitrary modules, not just <a href="https://pytorch.org/docs/stable/generated/torch.nn.Linear.html" rel="nofollow">nn.Linear</a> modules.</li> <li>Targeted support for specific modules by name or class.</li></ul> <p data-svelte-h="svelte-zg8sal">Install compressed-tensors from <a href="https://pypi.org/project/compressed-tensors" rel="nofollow">PyPI</a> to get the latest stable release (recommended) or install it from source to get the latest features.</p> <div class="flex space-x-2 items-center my-1.5 mr-8 h-7 !pl-0 -mx-3 md:mx-0"><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd border-gray-800 bg-black dark:bg-gray-700 text-white">PyPI </div><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd text-gray-500 cursor-pointer opacity-90 hover:text-gray-700 dark:hover:text-gray-200 hover:shadow-sm">source code </div></div> <div class="language-select"><div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->pip install compressed-tensors<!-- HTML_TAG_END --></pre></div> </div> <p data-svelte-h="svelte-o4tfur">Search using the compressed-tensors <a href="https://huggingface.co/models?other=compressed-tensors" rel="nofollow">tag</a> to find a compatible model on the Hugging Face Hub.</p> <p data-svelte-h="svelte-97790q">Only models that have already been quantized can be loaded at the moment, and once a model is loaded, it cannot be saved. To quantize a model into the compressed-tensors format, see <a href="https://github.com/vllm-project/llm-compressor" rel="nofollow">llm-compressor</a>. Alternatively, models can be created independently and serizlied with a compressed-tensors config.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoModelForCausalLM

	ct_model = AutoModelForCausalLM.from_pretrained(<span class="hljs-string">"nm-testing/Meta-Llama-3.1-8B-Instruct-FP8-hf"</span>, device_map=<span class="hljs-string">"auto"</span>)

	<span class="hljs-comment"># measure memory usage</span>
	mem_params = <span class="hljs-built_in">sum</span>([param.nelement()*param.element_size() <span class="hljs-keyword">for</span> param <span class="hljs-keyword">in</span> ct_model.parameters()])
	<span class="hljs-built_in">print</span>(<span class="hljs-string">f"<span class="hljs-subst">{mem_params/<span class="hljs-number">2</span>**<span class="hljs-number">30</span>:<span class="hljs-number">.4</span>f}</span> GB"</span>)
	<span class="hljs-comment"># 8.4575 GB</span><!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="model-checkpoint" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#model-checkpoint"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Model checkpoint</span></h2> <p data-svelte-h="svelte-2k8zy">Compressed-tensor models are defined through its configuration entry. The following example is taken from the <a href="https://huggingface.co/nm-testing/Meta-Llama-3.1-8B-Instruct-FP8-hf/blob/main/config.json" rel="nofollow">nm-testing/Meta-Llama-3.1-8B-Instruct-FP8-hf</a> <code>config.json</code> file.</p> <p data-svelte-h="svelte-fipnho">There are a lot of entries to allow for flexible expression both during and after compression, but the entries for loading and inference can be simplified to focus on just a few key entries.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-attr">"quantization_config"</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
	<span class="hljs-attr">"config_groups"</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
	<span class="hljs-attr">"group_0"</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
	<span class="hljs-attr">"input_activations"</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
	<span class="hljs-attr">"num_bits"</span><span class="hljs-punctuation">:</span> <span class="hljs-number">8</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"strategy"</span><span class="hljs-punctuation">:</span> <span class="hljs-string">"tensor"</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"type"</span><span class="hljs-punctuation">:</span> <span class="hljs-string">"float"</span>
	<span class="hljs-punctuation">}</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"targets"</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">[</span><span class="hljs-string">"Linear"</span><span class="hljs-punctuation">]</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"weights"</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
	<span class="hljs-attr">"num_bits"</span><span class="hljs-punctuation">:</span> <span class="hljs-number">8</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"strategy"</span><span class="hljs-punctuation">:</span> <span class="hljs-string">"tensor"</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"type"</span><span class="hljs-punctuation">:</span> <span class="hljs-string">"float"</span>
	<span class="hljs-punctuation">}</span>
	<span class="hljs-punctuation">}</span>
	<span class="hljs-punctuation">}</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"format"</span><span class="hljs-punctuation">:</span> <span class="hljs-string">"naive-quantized"</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"ignore"</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">[</span><span class="hljs-string">"lm_head"</span><span class="hljs-punctuation">]</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"quant_method"</span><span class="hljs-punctuation">:</span> <span class="hljs-string">"compressed-tensors"</span><span class="hljs-punctuation">,</span>
	<span class="hljs-attr">"quantization_status"</span><span class="hljs-punctuation">:</span> <span class="hljs-string">"frozen"</span>
	<span class="hljs-punctuation">}</span><span class="hljs-punctuation">,</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-67wddu">The config file specifies the quantization of a config group (<code>group_0</code>), which includes weight and activation quantization to fp8 with a static per-tensor strategy. The <code>lm_head</code> module is unquantized as shown in the <code>ignore</code> key.</p> <p data-svelte-h="svelte-xjewi7">For a more detailed look at the model weights, use the <a href="https://huggingface.co/nm-testing/Meta-Llama-3.1-8B-Instruct-FP8-hf?show_file_info=model.safetensors.index.json" rel="nofollow">safetensors viewer</a> on the model card to see the quantized weights, input scale, and weight scale for all <a href="https://pytorch.org/docs/stable/generated/torch.nn.Linear.html" rel="nofollow">nn.Linear</a> modules.</p> <table data-svelte-h="svelte-1ol90rr"><thead><tr><th>Tensors</th> <th>Shape</th> <th>Precision</th></tr></thead> <tbody><tr><td>model.layers.0.input_layernorm.weight</td> <td>[4 096]</td> <td>BF16</td></tr> <tr><td>model.layers.0.mlp.down_proj.input_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.mlp.down_proj.weight</td> <td>[4 096, 14 336]</td> <td>F8_E4M3</td></tr> <tr><td>model.layers.0.mlp.down_proj.weight_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.mlp.gate_proj.input_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.mlp.gate_proj.weight</td> <td>[14 336, 4 096]</td> <td>F8_E4M3</td></tr> <tr><td>model.layers.0.mlp.gate_proj.weight_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.mlp.up_proj.input_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.mlp.up_proj.weight</td> <td>[14 336, 4 096]</td> <td>F8_E4M3</td></tr> <tr><td>model.layers.0.mlp.up_proj.weight_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.post_attention_layernorm.weight</td> <td>[4 096]</td> <td>BF16</td></tr> <tr><td>model.layers.0.self_attn.k_proj.input_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.self_attn.k_proj.weight</td> <td>[1 024, 4 096]</td> <td>F8_E4M3</td></tr> <tr><td>model.layers.0.self_attn.k_proj.weight_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.self_attn.o_proj.input_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.self_attn.o_proj.weight</td> <td>[4 096, 4 096]</td> <td>F8_E4M3</td></tr> <tr><td>model.layers.0.self_attn.o_proj.weight_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.self_attn.q_proj.input_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.self_attn.q_proj.weight</td> <td>[4 096, 4 096]</td> <td>F8_E4M3</td></tr> <tr><td>model.layers.0.self_attn.q_proj.weight_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.self_attn.v_proj.input_scale</td> <td>[1]</td> <td>BF16</td></tr> <tr><td>model.layers.0.self_attn.v_proj.weight</td> <td>[1 024, 4 096]</td> <td>F8_E4M3</td></tr> <tr><td>model.layers.0.self_attn.v_proj.weight_scale</td> <td>[1]</td> <td>BF16</td></tr></tbody></table> <p data-svelte-h="svelte-9js76q">When loading a compressed-tensors model with the <code>~quantizers.HFQuantizer</code> integration, all the <a href="https://pytorch.org/docs/stable/generated/torch.nn.Linear.html" rel="nofollow">nn.Linear</a> modules specified in the quantization config are replaced by <a href="https://github.com/neuralmagic/compressed-tensors/blob/975cb223b19fcac2b98a4271d17668462d4d6e1d/src/compressed_tensors/linear/compressed_linear.py#L30" rel="nofollow">CompressedLinear</a> modules that manage the compressed weights and forward pass for inference. The <code>lm_head</code> module is still kept as an unquantized nn.Linear module.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoModelForCausalLM

	ct_model = AutoModelForCausalLM.from_pretrained(<span class="hljs-string">"nm-testing/Meta-Llama-3.1-8B-Instruct-FP8-hf"</span>)
	<span class="hljs-built_in">print</span>(ct_model)
	<span class="hljs-string">"""
	LlamaForCausalLM(
	(model): LlamaModel(
	(embed_tokens): Embedding(128256, 4096)
	(layers): ModuleList(
	(0-31): 32 x LlamaDecoderLayer(
	(self_attn): LlamaSdpaAttention(
	(q_proj): CompressedLinear(
	in_features=4096, out_features=4096, bias=False
	(input_observer): MovingAverageMinMaxObserver()
	(weight_observer): MovingAverageMinMaxObserver()
	)
	(k_proj): CompressedLinear(
	in_features=4096, out_features=1024, bias=False
	(input_observer): MovingAverageMinMaxObserver()
	(weight_observer): MovingAverageMinMaxObserver()
	)
	(v_proj): CompressedLinear(
	in_features=4096, out_features=1024, bias=False
	(input_observer): MovingAverageMinMaxObserver()
	(weight_observer): MovingAverageMinMaxObserver()
	)
	(o_proj): CompressedLinear(
	in_features=4096, out_features=4096, bias=False
	(input_observer): MovingAverageMinMaxObserver()
	(weight_observer): MovingAverageMinMaxObserver()
	)
	(rotary_emb): LlamaRotaryEmbedding()
	)
	(mlp): LlamaMLP(
	(gate_proj): CompressedLinear(
	in_features=4096, out_features=14336, bias=False
	(input_observer): MovingAverageMinMaxObserver()
	(weight_observer): MovingAverageMinMaxObserver()
	)
	(up_proj): CompressedLinear(
	in_features=4096, out_features=14336, bias=False
	(input_observer): MovingAverageMinMaxObserver()
	(weight_observer): MovingAverageMinMaxObserver()
	)
	(down_proj): CompressedLinear(
	in_features=14336, out_features=4096, bias=False
	(input_observer): MovingAverageMinMaxObserver()
	(weight_observer): MovingAverageMinMaxObserver()
	)
	(act_fn): SiLU()
	)
	(input_layernorm): LlamaRMSNorm((4096,), eps=1e-05)
	(post_attention_layernorm): LlamaRMSNorm((4096,), eps=1e-05)
	)
	)
	(norm): LlamaRMSNorm((4096,), eps=1e-05)
	(rotary_emb): LlamaRotaryEmbedding()
	)
	(lm_head): Linear(in_features=4096, out_features=128256, bias=False)
	)
	"""</span><!-- HTML_TAG_END --></pre></div> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/transformers/blob/main/docs/source/en/quantization/compressed_tensors.md" target="_blank"><svg class="mr-1" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M31,16l-7,7l-1.41-1.41L28.17,16l-5.58-5.59L24,9l7,7z"></path><path d="M1,16l7-7l1.41,1.41L3.83,16l5.58,5.59L8,23l-7-7z"></path><path d="M12.419,25.484L17.639,6.552l1.932,0.518L14.351,26.002z"></path></svg> <span data-svelte-h="svelte-zjs2n5"><span class="underline">Update</span> on GitHub</span></a> <p></p>

	<script>
	{
	__sveltekit_16tnnm8 = {
	assets: "/docs/transformers/pr_33892/en",
	base: "/docs/transformers/pr_33892/en",
	env: {}
	};

	const element = document.currentScript.parentElement;

	const data = [null,null];

	Promise.all([
	import("/docs/transformers/pr_33892/en/_app/immutable/entry/start.b2c4257a.js"),
	import("/docs/transformers/pr_33892/en/_app/immutable/entry/app.05ef1f97.js")
	]).then(([kit, app]) => {
	kit.start(app, element, {
	node_ids: [0, 526],
	data,
	form: null,
	error: null
	});
	});
	}
	</script>

Xet Storage Details

Size:: 26.7 kB
Xet hash:: 76705fe92f7044e6aa2871c17c644e028349b2a2b9c52c121ab016a929135c25

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.