Buckets:

hf-doc-build
/

doc

Files

xet

hf-doc-build/doc / peft /main /en /task_guides /image_classification_lora.html

rtrm

4 months ago

download

raw

73.5 kB

	<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"Image classification using LoRA","local":"image-classification-using-lora","sections":[{"title":"Install dependencies","local":"install-dependencies","sections":[],"depth":2},{"title":"Authenticate to share your model","local":"authenticate-to-share-your-model","sections":[],"depth":2},{"title":"Select a model checkpoint to fine-tune","local":"select-a-model-checkpoint-to-fine-tune","sections":[],"depth":2},{"title":"Load a dataset","local":"load-a-dataset","sections":[],"depth":2},{"title":"Dataset preparation","local":"dataset-preparation","sections":[],"depth":2},{"title":"Load and prepare a model","local":"load-and-prepare-a-model","sections":[],"depth":2},{"title":"Define training arguments","local":"define-training-arguments","sections":[],"depth":2},{"title":"Prepare evaluation metric","local":"prepare-evaluation-metric","sections":[],"depth":2},{"title":"Define collation function","local":"define-collation-function","sections":[],"depth":2},{"title":"Train and evaluate","local":"train-and-evaluate","sections":[],"depth":2},{"title":"Share your model and run inference","local":"share-your-model-and-run-inference","sections":[],"depth":2}],"depth":1}">
	<link href="/docs/peft/main/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/entry/start.f1bde2f6.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/chunks/scheduler.d627b047.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/chunks/singletons.eea5c746.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/chunks/index.a57a1c33.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/chunks/paths.e43091d4.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/entry/app.182b7f64.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/chunks/index.d48c4817.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/nodes/0.8a16be44.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/chunks/each.e59479a4.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/nodes/35.c42b6162.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/chunks/CodeBlock.5da89496.js">
	<link rel="modulepreload" href="/docs/peft/main/en/_app/immutable/chunks/Heading.47e562a9.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{"title":"Image classification using LoRA","local":"image-classification-using-lora","sections":[{"title":"Install dependencies","local":"install-dependencies","sections":[],"depth":2},{"title":"Authenticate to share your model","local":"authenticate-to-share-your-model","sections":[],"depth":2},{"title":"Select a model checkpoint to fine-tune","local":"select-a-model-checkpoint-to-fine-tune","sections":[],"depth":2},{"title":"Load a dataset","local":"load-a-dataset","sections":[],"depth":2},{"title":"Dataset preparation","local":"dataset-preparation","sections":[],"depth":2},{"title":"Load and prepare a model","local":"load-and-prepare-a-model","sections":[],"depth":2},{"title":"Define training arguments","local":"define-training-arguments","sections":[],"depth":2},{"title":"Prepare evaluation metric","local":"prepare-evaluation-metric","sections":[],"depth":2},{"title":"Define collation function","local":"define-collation-function","sections":[],"depth":2},{"title":"Train and evaluate","local":"train-and-evaluate","sections":[],"depth":2},{"title":"Share your model and run inference","local":"share-your-model-and-run-inference","sections":[],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <h1 class="relative group"><a id="image-classification-using-lora" 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="#image-classification-using-lora"><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>Image classification using LoRA</span></h1> <p data-svelte-h="svelte-1u144i3">This guide demonstrates how to use LoRA, a low-rank approximation technique, to fine-tune an image classification model.
	By using LoRA from 🤗 PEFT, we can reduce the number of trainable parameters in the model to only 0.77% of the original.</p> <p data-svelte-h="svelte-1q529js">LoRA achieves this reduction by adding low-rank “update matrices” to specific blocks of the model, such as the attention
	blocks. During fine-tuning, only these matrices are trained, while the original model parameters are left unchanged.
	At inference time, the update matrices are merged with the original model parameters to produce the final classification result.</p> <p data-svelte-h="svelte-e5fkf5">For more information on LoRA, please refer to the <a href="https://arxiv.org/abs/2106.09685" rel="nofollow">original LoRA paper</a>.</p> <h2 class="relative group"><a id="install-dependencies" 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="#install-dependencies"><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>Install dependencies</span></h2> <p data-svelte-h="svelte-1hz22x2">Install the libraries required for model training:</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 -->!pip install transformers accelerate evaluate datasets peft -q<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1stn34r">Check the versions of all required libraries to make sure you are up to date:</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">import</span> transformers
	<span class="hljs-keyword">import</span> accelerate
	<span class="hljs-keyword">import</span> peft

	<span class="hljs-built_in">print</span>(<span class="hljs-string">f"Transformers version: <span class="hljs-subst">{transformers.__version__}</span>"</span>)
	<span class="hljs-built_in">print</span>(<span class="hljs-string">f"Accelerate version: <span class="hljs-subst">{accelerate.__version__}</span>"</span>)
	<span class="hljs-built_in">print</span>(<span class="hljs-string">f"PEFT version: <span class="hljs-subst">{peft.__version__}</span>"</span>)
	<span class="hljs-string">"Transformers version: 4.27.4"</span>
	<span class="hljs-string">"Accelerate version: 0.18.0"</span>
	<span class="hljs-string">"PEFT version: 0.2.0"</span><!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="authenticate-to-share-your-model" 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="#authenticate-to-share-your-model"><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>Authenticate to share your model</span></h2> <p data-svelte-h="svelte-18wc95j">To share the fine-tuned model at the end of the training with the community, authenticate using your 🤗 token.
	You can obtain your token from your <a href="https://huggingface.co/settings/token" rel="nofollow">account settings</a>.</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> huggingface_hub <span class="hljs-keyword">import</span> notebook_login

	notebook_login()<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="select-a-model-checkpoint-to-fine-tune" 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="#select-a-model-checkpoint-to-fine-tune"><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>Select a model checkpoint to fine-tune</span></h2> <p data-svelte-h="svelte-1txdixu">Choose a model checkpoint from any of the model architectures supported for <a href="https://huggingface.co/models?pipeline_tag=image-classification&sort=downloads" rel="nofollow">image classification</a>. When in doubt, refer to
	the <a href="https://huggingface.co/docs/transformers/v4.27.2/en/tasks/image_classification" rel="nofollow">image classification task guide</a> in
	🤗 Transformers documentation.</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 -->model_checkpoint = <span class="hljs-string">"google/vit-base-patch16-224-in21k"</span><!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="load-a-dataset" 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="#load-a-dataset"><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>Load a dataset</span></h2> <p data-svelte-h="svelte-nfbivu">To keep this example’s runtime short, let’s only load the first 5000 instances from the training set of the <a href="https://huggingface.co/datasets/food101" rel="nofollow">Food-101 dataset</a>:</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> datasets <span class="hljs-keyword">import</span> load_dataset

	dataset = load_dataset(<span class="hljs-string">"food101"</span>, split=<span class="hljs-string">"train[:5000]"</span>)<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="dataset-preparation" 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="#dataset-preparation"><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>Dataset preparation</span></h2> <p data-svelte-h="svelte-rpv4zv">To prepare the dataset for training and evaluation, create <code>label2id</code> and <code>id2label</code> dictionaries. These will come in
	handy when performing inference and for metadata information:</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 -->labels = dataset.features[<span class="hljs-string">"label"</span>].names
	label2id, id2label = <span class="hljs-built_in">dict</span>(), <span class="hljs-built_in">dict</span>()
	<span class="hljs-keyword">for</span> i, label <span class="hljs-keyword">in</span> <span class="hljs-built_in">enumerate</span>(labels):
	label2id[label] = i
	id2label[i] = label

	id2label[<span class="hljs-number">2</span>]
	<span class="hljs-string">"baklava"</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1jp78mf">Next, load the image processor of the model you’re fine-tuning:</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> AutoImageProcessor

	image_processor = AutoImageProcessor.from_pretrained(model_checkpoint)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1b2m298">The <code>image_processor</code> contains useful information on which size the training and evaluation images should be resized
	to, as well as values that should be used to normalize the pixel values. Using the <code>image_processor</code>, prepare transformation
	functions for the datasets. These functions will include data augmentation and pixel scaling:</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> torchvision.transforms <span class="hljs-keyword">import</span> (
	CenterCrop,
	Compose,
	Normalize,
	RandomHorizontalFlip,
	RandomResizedCrop,
	Resize,
	ToTensor,
	)

	normalize = Normalize(mean=image_processor.image_mean, std=image_processor.image_std)
	train_transforms = Compose(
	[
	RandomResizedCrop(image_processor.size[<span class="hljs-string">"height"</span>]),
	RandomHorizontalFlip(),
	ToTensor(),
	normalize,
	]
	)

	val_transforms = Compose(
	[
	Resize(image_processor.size[<span class="hljs-string">"height"</span>]),
	CenterCrop(image_processor.size[<span class="hljs-string">"height"</span>]),
	ToTensor(),
	normalize,
	]
	)


	<span class="hljs-keyword">def</span> <span class="hljs-title function_">preprocess_train</span>(<span class="hljs-params">example_batch</span>):
	<span class="hljs-string">"""Apply train_transforms across a batch."""</span>
	example_batch[<span class="hljs-string">"pixel_values"</span>] = [train_transforms(image.convert(<span class="hljs-string">"RGB"</span>)) <span class="hljs-keyword">for</span> image <span class="hljs-keyword">in</span> example_batch[<span class="hljs-string">"image"</span>]]
	<span class="hljs-keyword">return</span> example_batch


	<span class="hljs-keyword">def</span> <span class="hljs-title function_">preprocess_val</span>(<span class="hljs-params">example_batch</span>):
	<span class="hljs-string">"""Apply val_transforms across a batch."""</span>
	example_batch[<span class="hljs-string">"pixel_values"</span>] = [val_transforms(image.convert(<span class="hljs-string">"RGB"</span>)) <span class="hljs-keyword">for</span> image <span class="hljs-keyword">in</span> example_batch[<span class="hljs-string">"image"</span>]]
	<span class="hljs-keyword">return</span> example_batch<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-v3a4as">Split the dataset into training and validation sets:</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 -->splits = dataset.train_test_split(test_size=<span class="hljs-number">0.1</span>)
	train_ds = splits[<span class="hljs-string">"train"</span>]
	val_ds = splits[<span class="hljs-string">"test"</span>]<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-k5a4bq">Finally, set the transformation functions for the datasets accordingly:</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 -->train_ds.set_transform(preprocess_train)
	val_ds.set_transform(preprocess_val)<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="load-and-prepare-a-model" 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="#load-and-prepare-a-model"><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>Load and prepare a model</span></h2> <p data-svelte-h="svelte-lmmzq">Before loading the model, let’s define a helper function to check the total number of parameters a model has, as well
	as how many of them are trainable.</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">def</span> <span class="hljs-title function_">print_trainable_parameters</span>(<span class="hljs-params">model</span>):
	trainable_params = <span class="hljs-number">0</span>
	all_param = <span class="hljs-number">0</span>
	<span class="hljs-keyword">for</span> _, param <span class="hljs-keyword">in</span> model.named_parameters():
	all_param += param.numel()
	<span class="hljs-keyword">if</span> param.requires_grad:
	trainable_params += param.numel()
	<span class="hljs-built_in">print</span>(
	<span class="hljs-string">f"trainable params: <span class="hljs-subst">{trainable_params}</span> \|\| all params: <span class="hljs-subst">{all_param}</span> \|\| trainable%: <span class="hljs-subst">{<span class="hljs-number">100</span> * trainable_params / all_param:<span class="hljs-number">.2</span>f}</span>"</span>
	)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1sp5d8a">It’s important to initialize the original model correctly as it will be used as a base to create the <code>PeftModel</code> you’ll
	actually fine-tune. Specify the <code>label2id</code> and <code>id2label</code> so that <a href="https://huggingface.co/docs/transformers/main/en/model_doc/auto#transformers.AutoModelForImageClassification" rel="nofollow">AutoModelForImageClassification</a> can append a classification
	head to the underlying model, adapted for this dataset. You should see the following output:</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">Some</span> weights <span class="hljs-keyword">of</span> ViTForImageClassification were <span class="hljs-keyword">not</span> initialized <span class="hljs-keyword">from</span> the model <span class="hljs-keyword">checkpoint</span> at google/vit-base-patch16<span class="hljs-number">-224</span>-in21k <span class="hljs-keyword">and</span> are newly initialized: [<span class="hljs-string">'classifier.weight'</span>, <span class="hljs-string">'classifier.bias'</span>]<!-- HTML_TAG_END --></pre></div> <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> AutoModelForImageClassification, TrainingArguments, Trainer

	model = AutoModelForImageClassification.from_pretrained(
	model_checkpoint,
	label2id=label2id,
	id2label=id2label,
	ignore_mismatched_sizes=<span class="hljs-literal">True</span>, <span class="hljs-comment"># provide this in case you're planning to fine-tune an already fine-tuned checkpoint</span>
	)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1ez48do">Before creating a <code>PeftModel</code>, you can check the number of trainable parameters in the original model:</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 -->print_trainable_parameters(model)
	<span class="hljs-string">"trainable params: 85876325 \|\| all params: 85876325 \|\| trainable%: 100.00"</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-wkppti">Next, use <code>get_peft_model</code> to wrap the base model so that “update” matrices are added to the respective places.</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> peft <span class="hljs-keyword">import</span> LoraConfig, get_peft_model

	config = LoraConfig(
	r=<span class="hljs-number">16</span>,
	lora_alpha=<span class="hljs-number">16</span>,
	target_modules=[<span class="hljs-string">"query"</span>, <span class="hljs-string">"value"</span>],
	lora_dropout=<span class="hljs-number">0.1</span>,
	bias=<span class="hljs-string">"none"</span>,
	modules_to_save=[<span class="hljs-string">"classifier"</span>],
	)
	lora_model = get_peft_model(model, config)
	print_trainable_parameters(lora_model)
	<span class="hljs-string">"trainable params: 667493 \|\| all params: 86466149 \|\| trainable%: 0.77"</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-q6mejx">Let’s unpack what’s going on here.
	To use LoRA, you need to specify the target modules in <code>LoraConfig</code> so that <code>get_peft_model()</code> knows which modules
	inside our model need to be amended with LoRA matrices. In this example, we’re only interested in targeting the query and
	value matrices of the attention blocks of the base model. Since the parameters corresponding to these matrices are “named”
	“query” and “value” respectively, we specify them accordingly in the <code>target_modules</code> argument of <code>LoraConfig</code>.</p> <p data-svelte-h="svelte-7y27bh">We also specify <code>modules_to_save</code>. After wrapping the base model with <code>get_peft_model()</code> along with the <code>config</code>, we get
	a new model where only the LoRA parameters are trainable (so-called “update matrices”) while the pre-trained parameters
	are kept frozen. However, we want the classifier parameters to be trained too when fine-tuning the base model on our
	custom dataset. To ensure that the classifier parameters are also trained, we specify <code>modules_to_save</code>. This also
	ensures that these modules are serialized alongside the LoRA trainable parameters when using utilities like <code>save_pretrained()</code>
	and <code>push_to_hub()</code>.</p> <p data-svelte-h="svelte-j5msxy">Here’s what the other parameters mean:</p> <ul data-svelte-h="svelte-1m2rv79"><li><code>r</code>: The dimension used by the LoRA update matrices.</li> <li><code>alpha</code>: Scaling factor.</li> <li><code>bias</code>: Specifies if the <code>bias</code> parameters should be trained. <code>None</code> denotes none of the <code>bias</code> parameters will be trained.</li></ul> <p data-svelte-h="svelte-v2eh0"><code>r</code> and <code>alpha</code> together control the total number of final trainable parameters when using LoRA, giving you the flexibility
	to balance a trade-off between end performance and compute efficiency.</p> <p data-svelte-h="svelte-bwxiph">By looking at the number of trainable parameters, you can see how many parameters we’re actually training. Since the goal is
	to achieve parameter-efficient fine-tuning, you should expect to see fewer trainable parameters in the <code>lora_model</code>
	in comparison to the original model, which is indeed the case here.</p> <h2 class="relative group"><a id="define-training-arguments" 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="#define-training-arguments"><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>Define training arguments</span></h2> <p data-svelte-h="svelte-132w99p">For model fine-tuning, use <a href="https://huggingface.co/docs/transformers/main/en/main_classes/trainer#transformers.Trainer" rel="nofollow">Trainer</a>. It accepts
	several arguments which you can wrap using <a href="https://huggingface.co/docs/transformers/main/en/main_classes/trainer#transformers.TrainingArguments" rel="nofollow">TrainingArguments</a>.</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> TrainingArguments, Trainer


	model_name = model_checkpoint.split(<span class="hljs-string">"/"</span>)[-<span class="hljs-number">1</span>]
	batch_size = <span class="hljs-number">128</span>

	args = TrainingArguments(
	<span class="hljs-string">f"<span class="hljs-subst">{model_name}</span>-finetuned-lora-food101"</span>,
	remove_unused_columns=<span class="hljs-literal">False</span>,
	evaluation_strategy=<span class="hljs-string">"epoch"</span>,
	save_strategy=<span class="hljs-string">"epoch"</span>,
	learning_rate=<span class="hljs-number">5e-3</span>,
	per_device_train_batch_size=batch_size,
	gradient_accumulation_steps=<span class="hljs-number">4</span>,
	per_device_eval_batch_size=batch_size,
	fp16=<span class="hljs-literal">True</span>,
	num_train_epochs=<span class="hljs-number">5</span>,
	logging_steps=<span class="hljs-number">10</span>,
	load_best_model_at_end=<span class="hljs-literal">True</span>,
	metric_for_best_model=<span class="hljs-string">"accuracy"</span>,
	push_to_hub=<span class="hljs-literal">True</span>,
	label_names=[<span class="hljs-string">"labels"</span>],
	)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1b1k1gd">Compared to non-PEFT methods, you can use a larger batch size since there are fewer parameters to train.
	You can also set a larger learning rate than the normal (1e-5 for example).</p> <p data-svelte-h="svelte-15s8rfp">This can potentially also reduce the need to conduct expensive hyperparameter tuning experiments.</p> <h2 class="relative group"><a id="prepare-evaluation-metric" 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="#prepare-evaluation-metric"><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>Prepare evaluation metric</span></h2> <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">import</span> numpy <span class="hljs-keyword">as</span> np
	<span class="hljs-keyword">import</span> evaluate

	metric = evaluate.load(<span class="hljs-string">"accuracy"</span>)


	<span class="hljs-keyword">def</span> <span class="hljs-title function_">compute_metrics</span>(<span class="hljs-params">eval_pred</span>):
	<span class="hljs-string">"""Computes accuracy on a batch of predictions"""</span>
	predictions = np.argmax(eval_pred.predictions, axis=<span class="hljs-number">1</span>)
	<span class="hljs-keyword">return</span> metric.compute(predictions=predictions, references=eval_pred.label_ids)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-csgsx6">The <code>compute_metrics</code> function takes a named tuple as input: <code>predictions</code>, which are the logits of the model as Numpy arrays,
	and <code>label_ids</code>, which are the ground-truth labels as Numpy arrays.</p> <h2 class="relative group"><a id="define-collation-function" 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="#define-collation-function"><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>Define collation function</span></h2> <p data-svelte-h="svelte-vp0cw5">A collation function is used by <a href="https://huggingface.co/docs/transformers/main/en/main_classes/trainer#transformers.Trainer" rel="nofollow">Trainer</a> to gather a batch of training and evaluation examples and prepare them in a
	format that is acceptable by the underlying model.</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">import</span> torch


	<span class="hljs-keyword">def</span> <span class="hljs-title function_">collate_fn</span>(<span class="hljs-params">examples</span>):
	pixel_values = torch.stack([example[<span class="hljs-string">"pixel_values"</span>] <span class="hljs-keyword">for</span> example <span class="hljs-keyword">in</span> examples])
	labels = torch.tensor([example[<span class="hljs-string">"label"</span>] <span class="hljs-keyword">for</span> example <span class="hljs-keyword">in</span> examples])
	<span class="hljs-keyword">return</span> {<span class="hljs-string">"pixel_values"</span>: pixel_values, <span class="hljs-string">"labels"</span>: labels}<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="train-and-evaluate" 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="#train-and-evaluate"><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>Train and evaluate</span></h2> <p data-svelte-h="svelte-ewvpbn">Bring everything together - model, training arguments, data, collation function, etc. Then, start the training!</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 -->trainer = Trainer(
	lora_model,
	args,
	train_dataset=train_ds,
	eval_dataset=val_ds,
	tokenizer=image_processor,
	compute_metrics=compute_metrics,
	data_collator=collate_fn,
	)
	train_results = trainer.train()<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-156715e">In just a few minutes, the fine-tuned model shows 96% validation accuracy even on this small
	subset of the training dataset.</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 -->trainer.evaluate(val_ds)
	{
	<span class="hljs-string">"eval_loss"</span>: <span class="hljs-number">0.14475855231285095</span>,
	<span class="hljs-string">"eval_accuracy"</span>: <span class="hljs-number">0.96</span>,
	<span class="hljs-string">"eval_runtime"</span>: <span class="hljs-number">3.5725</span>,
	<span class="hljs-string">"eval_samples_per_second"</span>: <span class="hljs-number">139.958</span>,
	<span class="hljs-string">"eval_steps_per_second"</span>: <span class="hljs-number">1.12</span>,
	<span class="hljs-string">"epoch"</span>: <span class="hljs-number">5.0</span>,
	}<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="share-your-model-and-run-inference" 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="#share-your-model-and-run-inference"><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>Share your model and run inference</span></h2> <p data-svelte-h="svelte-1nwbsod">Once the fine-tuning is done, share the LoRA parameters with the community like so:</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 -->repo_name = <span class="hljs-string">f"sayakpaul/<span class="hljs-subst">{model_name}</span>-finetuned-lora-food101"</span>
	lora_model.push_to_hub(repo_name)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1yy7agj">When calling <a href="https://huggingface.co/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel.push_to_hub" rel="nofollow">push_to_hub</a> on the <code>lora_model</code>, only the LoRA parameters along with any modules specified in <code>modules_to_save</code>
	are saved. Take a look at the <a href="https://huggingface.co/sayakpaul/vit-base-patch16-224-in21k-finetuned-lora-food101/blob/main/adapter_model.bin" rel="nofollow">trained LoRA parameters</a>.
	You’ll see that it’s only 2.6 MB! This greatly helps with portability, especially when using a very large model to fine-tune (such as <a href="https://huggingface.co/bigscience/bloom" rel="nofollow">BLOOM</a>).</p> <p data-svelte-h="svelte-1sywq05">Next, let’s see how to load the LoRA updated parameters along with our base model for inference. When you wrap a base model
	with <code>PeftModel</code>, modifications are done <em>in-place</em>. To mitigate any concerns that might stem from in-place modifications,
	initialize the base model just like you did earlier and construct the inference model.</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> peft <span class="hljs-keyword">import</span> PeftConfig, PeftModel


	config = PeftConfig.from_pretrained(repo_name)
	model = AutoModelForImageClassification.from_pretrained(
	config.base_model_name_or_path,
	label2id=label2id,
	id2label=id2label,
	ignore_mismatched_sizes=<span class="hljs-literal">True</span>, <span class="hljs-comment"># provide this in case you're planning to fine-tune an already fine-tuned checkpoint</span>
	)
	<span class="hljs-comment"># Load the LoRA model</span>
	inference_model = PeftModel.from_pretrained(model, repo_name)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-143jykz">Let’s now fetch an example image for inference.</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> PIL <span class="hljs-keyword">import</span> Image
	<span class="hljs-keyword">import</span> requests

	url = <span class="hljs-string">"https://huggingface.co/datasets/sayakpaul/sample-datasets/resolve/main/beignets.jpeg"</span>
	image = Image.<span class="hljs-built_in">open</span>(requests.get(url, stream=<span class="hljs-literal">True</span>).raw)
	image<!-- HTML_TAG_END --></pre></div> <div class="flex justify-center" data-svelte-h="svelte-1y4h4v8"><img src="https://huggingface.co/datasets/sayakpaul/sample-datasets/resolve/main/beignets.jpeg" alt="image of beignets"></div> <p data-svelte-h="svelte-snna81">First, instantiate an <code>image_processor</code> from the underlying model repo.</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 -->image_processor = AutoImageProcessor.from_pretrained(repo_name)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-oy7q59">Then, prepare the example for inference.</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 -->encoding = image_processor(image.convert(<span class="hljs-string">"RGB"</span>), return_tensors=<span class="hljs-string">"pt"</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1u4plry">Finally, run inference!</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">with</span> torch.no_grad():
	outputs = inference_model(**encoding)
	logits = outputs.logits

	predicted_class_idx = logits.argmax(-<span class="hljs-number">1</span>).item()
	<span class="hljs-built_in">print</span>(<span class="hljs-string">"Predicted class:"</span>, inference_model.config.id2label[predicted_class_idx])
	<span class="hljs-string">"Predicted class: beignets"</span><!-- HTML_TAG_END --></pre></div> <p></p>

	<script>
	{
	__sveltekit_c8wad6 = {
	assets: "/docs/peft/main/en",
	base: "/docs/peft/main/en",
	env: {}
	};

	const element = document.currentScript.parentElement;

	const data = [null,null];

	Promise.all([
	import("/docs/peft/main/en/_app/immutable/entry/start.f1bde2f6.js"),
	import("/docs/peft/main/en/_app/immutable/entry/app.182b7f64.js")
	]).then(([kit, app]) => {
	kit.start(app, element, {
	node_ids: [0, 35],
	data,
	form: null,
	error: null
	});
	});
	}
	</script>

Xet Storage Details

Size:: 73.5 kB
Xet hash:: bbc0702c668168db5d4af95e98215f9647c44119c486323c17a70ff759cee3e6

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