Buckets:
| <meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"Train and deploy Hugging Face on Amazon SageMaker","local":"train-and-deploy-hugging-face-on-amazon-sagemaker","sections":[{"title":"Installation and setup","local":"installation-and-setup","sections":[],"depth":2},{"title":"Preprocess","local":"preprocess","sections":[],"depth":2},{"title":"Upload dataset to S3 bucket","local":"upload-dataset-to-s3-bucket","sections":[],"depth":2},{"title":"Start a training job","local":"start-a-training-job","sections":[],"depth":2},{"title":"Deploy model","local":"deploy-model","sections":[],"depth":2},{"title":"What’s next?","local":"whats-next","sections":[],"depth":2}],"depth":1}"> | |
| <link href="/docs/sagemaker/pr_1645/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/entry/start.f559cddd.js"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/chunks/scheduler.389d799c.js"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/chunks/singletons.d40c3c49.js"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/chunks/paths.9e868e10.js"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/entry/app.d4de874a.js"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/chunks/index.8f81d18f.js"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/nodes/0.acb1a48c.js"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/nodes/2.a0caa31f.js"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/chunks/CodeBlock.3845caa1.js"> | |
| <link rel="modulepreload" href="/docs/sagemaker/pr_1645/en/_app/immutable/chunks/EditOnGithub.33306dfe.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{"title":"Train and deploy Hugging Face on Amazon SageMaker","local":"train-and-deploy-hugging-face-on-amazon-sagemaker","sections":[{"title":"Installation and setup","local":"installation-and-setup","sections":[],"depth":2},{"title":"Preprocess","local":"preprocess","sections":[],"depth":2},{"title":"Upload dataset to S3 bucket","local":"upload-dataset-to-s3-bucket","sections":[],"depth":2},{"title":"Start a training job","local":"start-a-training-job","sections":[],"depth":2},{"title":"Deploy model","local":"deploy-model","sections":[],"depth":2},{"title":"What’s next?","local":"whats-next","sections":[],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <h1 class="relative group"><a id="train-and-deploy-hugging-face-on-amazon-sagemaker" 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-deploy-hugging-face-on-amazon-sagemaker"><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 deploy Hugging Face on Amazon SageMaker</span></h1> <p data-svelte-h="svelte-n7q5fy">The get started guide will show you how to quickly use Hugging Face on Amazon SageMaker. Learn how to fine-tune and deploy a pretrained 🤗 Transformers model on SageMaker for a binary text classification task.</p> <p data-svelte-h="svelte-uamv5h">💡 If you are new to Hugging Face, we recommend first reading the 🤗 Transformers <a href="https://huggingface.co/docs/transformers/quicktour" rel="nofollow">quick tour</a>.</p> <iframe width="560" height="315" src="https://www.youtube.com/embed/pYqjCzoyWyo" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen=""></iframe> <p data-svelte-h="svelte-rt6seb">📓 Open the <a href="https://github.com/huggingface/notebooks/blob/main/sagemaker/01_getting_started_pytorch/sagemaker-notebook.ipynb" rel="nofollow">agemaker-notebook.ipynb file</a> to follow along!</p> <h2 class="relative group"><a id="installation-and-setup" 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="#installation-and-setup"><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>Installation and setup</span></h2> <p data-svelte-h="svelte-10ldwp9">Get started by installing the necessary Hugging Face libraries and SageMaker. You will also need to install <a href="https://pytorch.org/get-started/locally/" rel="nofollow">PyTorch</a> and <a href="https://www.tensorflow.org/install/pip#tensorflow-2-packages-are-available" rel="nofollow">TensorFlow</a> if you don’t already have it installed.</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 <span class="hljs-string">"sagemaker>=2.140.0"</span> <span class="hljs-string">"transformers==4.26.1"</span> <span class="hljs-string">"datasets[s3]==2.10.1"</span> --upgrade<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-yozy2b">If you want to run this example in <a href="https://docs.aws.amazon.com/sagemaker/latest/dg/studio.html" rel="nofollow">SageMaker Studio</a>, upgrade <a href="https://ipywidgets.readthedocs.io/en/latest/" rel="nofollow">ipywidgets</a> for the 🤗 Datasets library and restart the kernel:</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 -->%%capture | |
| <span class="hljs-keyword">import</span> IPython | |
| !conda install -c conda-forge ipywidgets -y | |
| IPython.Application.instance().kernel.do_shutdown(<span class="hljs-literal">True</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1sllt2f">Next, you should set up your environment: a SageMaker session and an S3 bucket. The S3 bucket will store data, models, and logs. You will need access to an <a href="https://docs.aws.amazon.com/sagemaker/latest/dg/sagemaker-roles.html" rel="nofollow">IAM execution role</a> with the required permissions.</p> <p data-svelte-h="svelte-tg9h77">If you are planning on using SageMaker in a local environment, you need to provide the <code>role</code> yourself. Learn more about how to set this up <a href="https://huggingface.co/docs/sagemaker/train#installation-and-setup" rel="nofollow">here</a>.</p> <p data-svelte-h="svelte-y9vgbx">⚠️ The execution role is only available when you run a notebook within SageMaker. If you try to run <code>get_execution_role</code> in a notebook not on SageMaker, you will get a region error.</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> sagemaker | |
| sess = sagemaker.Session() | |
| sagemaker_session_bucket = <span class="hljs-literal">None</span> | |
| <span class="hljs-keyword">if</span> sagemaker_session_bucket <span class="hljs-keyword">is</span> <span class="hljs-literal">None</span> <span class="hljs-keyword">and</span> sess <span class="hljs-keyword">is</span> <span class="hljs-keyword">not</span> <span class="hljs-literal">None</span>: | |
| sagemaker_session_bucket = sess.default_bucket() | |
| role = sagemaker.get_execution_role() | |
| sess = sagemaker.Session(default_bucket=sagemaker_session_bucket)<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="preprocess" 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="#preprocess"><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>Preprocess</span></h2> <p data-svelte-h="svelte-b8daui">The 🤗 Datasets library makes it easy to download and preprocess a dataset for training. Download and tokenize the <a href="https://huggingface.co/datasets/imdb" rel="nofollow">IMDb</a> 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 --><span class="hljs-keyword">from</span> datasets <span class="hljs-keyword">import</span> load_dataset | |
| <span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoTokenizer | |
| <span class="hljs-comment"># load dataset</span> | |
| train_dataset, test_dataset = load_dataset(<span class="hljs-string">"imdb"</span>, split=[<span class="hljs-string">"train"</span>, <span class="hljs-string">"test"</span>]) | |
| <span class="hljs-comment"># load tokenizer</span> | |
| tokenizer = AutoTokenizer.from_pretrained(<span class="hljs-string">"distilbert/distilbert-base-uncased"</span>) | |
| <span class="hljs-comment"># create tokenization function</span> | |
| <span class="hljs-keyword">def</span> <span class="hljs-title function_">tokenize</span>(<span class="hljs-params">batch</span>): | |
| <span class="hljs-keyword">return</span> tokenizer(batch[<span class="hljs-string">"text"</span>], padding=<span class="hljs-string">"max_length"</span>, truncation=<span class="hljs-literal">True</span>) | |
| <span class="hljs-comment"># tokenize train and test datasets</span> | |
| train_dataset = train_dataset.<span class="hljs-built_in">map</span>(tokenize, batched=<span class="hljs-literal">True</span>) | |
| test_dataset = test_dataset.<span class="hljs-built_in">map</span>(tokenize, batched=<span class="hljs-literal">True</span>) | |
| <span class="hljs-comment"># set dataset format for PyTorch</span> | |
| train_dataset = train_dataset.rename_column(<span class="hljs-string">"label"</span>, <span class="hljs-string">"labels"</span>) | |
| train_dataset.set_format(<span class="hljs-string">"torch"</span>, columns=[<span class="hljs-string">"input_ids"</span>, <span class="hljs-string">"attention_mask"</span>, <span class="hljs-string">"labels"</span>]) | |
| test_dataset = test_dataset.rename_column(<span class="hljs-string">"label"</span>, <span class="hljs-string">"labels"</span>) | |
| test_dataset.set_format(<span class="hljs-string">"torch"</span>, columns=[<span class="hljs-string">"input_ids"</span>, <span class="hljs-string">"attention_mask"</span>, <span class="hljs-string">"labels"</span>])<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="upload-dataset-to-s3-bucket" 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="#upload-dataset-to-s3-bucket"><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>Upload dataset to S3 bucket</span></h2> <p data-svelte-h="svelte-nm384r">Next, upload the preprocessed dataset to your S3 session bucket with 🤗 Datasets S3 <a href="https://huggingface.co/docs/datasets/filesystems.html" rel="nofollow">filesystem</a> implementation:</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-comment"># save train_dataset to s3</span> | |
| training_input_path = <span class="hljs-string">f's3://<span class="hljs-subst">{sess.default_bucket()}</span>/<span class="hljs-subst">{s3_prefix}</span>/train'</span> | |
| train_dataset.save_to_disk(training_input_path) | |
| <span class="hljs-comment"># save test_dataset to s3</span> | |
| test_input_path = <span class="hljs-string">f's3://<span class="hljs-subst">{sess.default_bucket()}</span>/<span class="hljs-subst">{s3_prefix}</span>/test'</span> | |
| test_dataset.save_to_disk(test_input_path)<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="start-a-training-job" 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="#start-a-training-job"><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>Start a training job</span></h2> <p data-svelte-h="svelte-16p4w97">Create a Hugging Face Estimator to handle end-to-end SageMaker training and deployment. The most important parameters to pay attention to are:</p> <ul data-svelte-h="svelte-1bnefx4"><li><code>entry_point</code> refers to the fine-tuning script which you can find in <a href="https://github.com/huggingface/notebooks/blob/main/sagemaker/01_getting_started_pytorch/scripts/train.py" rel="nofollow">train.py file</a>.</li> <li><code>instance_type</code> refers to the SageMaker instance that will be launched. Take a look <a href="https://aws.amazon.com/sagemaker/pricing/" rel="nofollow">here</a> for a complete list of instance types.</li> <li><code>hyperparameters</code> refers to the training hyperparameters the model will be fine-tuned with.</li></ul> <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> sagemaker.huggingface <span class="hljs-keyword">import</span> HuggingFace | |
| hyperparameters={ | |
| <span class="hljs-string">"epochs"</span>: <span class="hljs-number">1</span>, <span class="hljs-comment"># number of training epochs</span> | |
| <span class="hljs-string">"train_batch_size"</span>: <span class="hljs-number">32</span>, <span class="hljs-comment"># training batch size</span> | |
| <span class="hljs-string">"model_name"</span>:<span class="hljs-string">"distilbert/distilbert-base-uncased"</span> <span class="hljs-comment"># name of pretrained model</span> | |
| } | |
| huggingface_estimator = HuggingFace( | |
| entry_point=<span class="hljs-string">"train.py"</span>, <span class="hljs-comment"># fine-tuning script to use in training job</span> | |
| source_dir=<span class="hljs-string">"./scripts"</span>, <span class="hljs-comment"># directory where fine-tuning script is stored</span> | |
| instance_type=<span class="hljs-string">"ml.p3.2xlarge"</span>, <span class="hljs-comment"># instance type</span> | |
| instance_count=<span class="hljs-number">1</span>, <span class="hljs-comment"># number of instances</span> | |
| role=role, <span class="hljs-comment"># IAM role used in training job to acccess AWS resources (S3)</span> | |
| transformers_version=<span class="hljs-string">"4.26"</span>, <span class="hljs-comment"># Transformers version</span> | |
| pytorch_version=<span class="hljs-string">"1.13"</span>, <span class="hljs-comment"># PyTorch version</span> | |
| py_version=<span class="hljs-string">"py39"</span>, <span class="hljs-comment"># Python version</span> | |
| hyperparameters=hyperparameters <span class="hljs-comment"># hyperparameters to use in training job</span> | |
| )<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-b865q9">Begin training with one line of code:</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 -->huggingface_estimator.fit({<span class="hljs-string">"train"</span>: training_input_path, <span class="hljs-string">"test"</span>: test_input_path})<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="deploy-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="#deploy-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>Deploy model</span></h2> <p data-svelte-h="svelte-twh61n">Once the training job is complete, deploy your fine-tuned model by calling <code>deploy()</code> with the number of instances and instance type:</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 -->predictor = huggingface_estimator.deploy(initial_instance_count=<span class="hljs-number">1</span>,<span class="hljs-string">"ml.g4dn.xlarge"</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-19s4ndj">Call <code>predict()</code> on your data:</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 -->sentiment_input = {<span class="hljs-string">"inputs"</span>: <span class="hljs-string">"It feels like a curtain closing...there was an elegance in the way they moved toward conclusion. No fan is going to watch and feel short-changed."</span>} | |
| predictor.predict(sentiment_input)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-l180zc">After running your request, delete the endpoint:</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 -->predictor.delete_endpoint()<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="whats-next" 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="#whats-next"><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>What’s next?</span></h2> <p data-svelte-h="svelte-1rjk4za">Congratulations, you’ve just fine-tuned and deployed a pretrained 🤗 Transformers model on SageMaker! 🎉</p> <p data-svelte-h="svelte-1qzgnqe">For your next steps, keep reading our documentation for more details about training and deployment. There are many interesting features such as <a href="/docs/sagemaker/train#distributed-training">distributed training</a> and <a href="/docs/sagemaker/train#spot-instances">Spot instances</a>.</p> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/hub-docs/blob/main/docs/sagemaker/getting-started.md" target="_blank"><span data-svelte-h="svelte-1kd6by1"><</span> <span data-svelte-h="svelte-x0xyl0">></span> <span data-svelte-h="svelte-1dajgef"><span class="underline ml-1.5">Update</span> on GitHub</span></a> <p></p> | |
| <script> | |
| { | |
| __sveltekit_nrm40j = { | |
| assets: "/docs/sagemaker/pr_1645/en", | |
| base: "/docs/sagemaker/pr_1645/en", | |
| env: {} | |
| }; | |
| const element = document.currentScript.parentElement; | |
| const data = [null,null]; | |
| Promise.all([ | |
| import("/docs/sagemaker/pr_1645/en/_app/immutable/entry/start.f559cddd.js"), | |
| import("/docs/sagemaker/pr_1645/en/_app/immutable/entry/app.d4de874a.js") | |
| ]).then(([kit, app]) => { | |
| kit.start(app, element, { | |
| node_ids: [0, 2], | |
| data, | |
| form: null, | |
| error: null | |
| }); | |
| }); | |
| } | |
| </script> | |
Xet Storage Details
- Size:
- 35.6 kB
- Xet hash:
- 3100252072f3b356e680b143f81ccc19af06d48ffb37f8f415af30195a27b434
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.