Buckets:
| <meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"Ajuste fino supervisado","local":"ajuste-fino-supervisado","sections":[{"title":"1️⃣ Plantillas de chat","local":"1-plantillas-de-chat","sections":[],"depth":2},{"title":"2️⃣ Ajuste fino supervisado","local":"2-ajuste-fino-supervisado","sections":[],"depth":2},{"title":"3️⃣ Adaptación de bajo rango (LoRA)","local":"3-adaptación-de-bajo-rango-lora","sections":[],"depth":2},{"title":"4️⃣ Evaluación","local":"4-evaluación","sections":[],"depth":2},{"title":"Referencias","local":"referencias","sections":[],"depth":2}],"depth":1}"> | |
| <link href="/docs/course/pr_1213/es/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/entry/start.36d27295.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/chunks/scheduler.505acc25.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/chunks/singletons.6865fa96.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/chunks/index.001f95d5.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/chunks/paths.ec28c642.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/entry/app.3b43d7f3.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/chunks/preload-helper.8c2bab6b.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/chunks/index.e22abd30.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/nodes/0.e2c0ea78.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/chunks/each.e59479a4.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/nodes/21.052c4eaa.js"> | |
| <link rel="modulepreload" href="/docs/course/pr_1213/es/_app/immutable/chunks/MermaidChart.svelte_svelte_type_style_lang.a144e953.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{"title":"Ajuste fino supervisado","local":"ajuste-fino-supervisado","sections":[{"title":"1️⃣ Plantillas de chat","local":"1-plantillas-de-chat","sections":[],"depth":2},{"title":"2️⃣ Ajuste fino supervisado","local":"2-ajuste-fino-supervisado","sections":[],"depth":2},{"title":"3️⃣ Adaptación de bajo rango (LoRA)","local":"3-adaptación-de-bajo-rango-lora","sections":[],"depth":2},{"title":"4️⃣ Evaluación","local":"4-evaluación","sections":[],"depth":2},{"title":"Referencias","local":"referencias","sections":[],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <div class="items-center shrink-0 min-w-[100px] max-sm:min-w-[50px] justify-end ml-auto flex" style="float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"><div class="inline-flex rounded-md max-sm:rounded-sm"><button class="inline-flex items-center gap-1 h-7 max-sm:h-7 px-2 max-sm:px-1.5 text-sm font-medium text-gray-800 border border-r-0 rounded-l-md max-sm:rounded-l-sm border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-live="polite"><span class="inline-flex items-center justify-center rounded-md p-0.5 max-sm:p-0 hover:text-gray-800 dark:hover:text-gray-200"><svg class="sm:size-3.5 size-3" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg></span> <span>Copy page</span></button> <button class="inline-flex items-center justify-center w-6 max-sm:w-5 h-7 max-sm:h-7 disabled:pointer-events-none text-sm text-gray-500 hover:text-gray-700 dark:hover:text-white rounded-r-md max-sm:rounded-r-sm border border-l transition border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-haspopup="menu" aria-expanded="false" aria-label="Open copy menu"><svg class="transition-transform text-gray-400 overflow-visible sm:size-3.5 size-3 rotate-0" width="1em" height="1em" viewBox="0 0 12 7" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M1 1L6 6L11 1" stroke="currentColor"></path></svg></button></div> </div> <h1 class="relative group"><a id="ajuste-fino-supervisado" 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="#ajuste-fino-supervisado"><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>Ajuste fino supervisado</span></h1> <p data-svelte-h="svelte-jzh0zd">En la <a href="/course/chapter2/2">Sección 2 del Capítulo 2</a>, vimos que los modelos de lenguaje generativos se pueden ajustar para tareas concretas, como la resumición y la respuesta a preguntas. Sin embargo, hoy en día es mucho más habitual ajustar modelos de lenguaje para una amplia variedad de tareas al mismo tiempo; un método conocido como ajuste fino supervisado (SFT, por sus siglas en inglés). Este proceso ayuda a que los modelos sean más versátiles y capaces de manejar casos de uso diversos. La mayoría de los LLM con los que la gente interactúa en plataformas como ChatGPT han pasado por SFT para ser más útiles y estar mejor alineados con las preferencias humanas. Dividiremos este capítulo en cuatro secciones:</p> <h2 class="relative group"><a id="1-plantillas-de-chat" 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="#1-plantillas-de-chat"><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>1️⃣ Plantillas de chat</span></h2> <p data-svelte-h="svelte-b9fjol">Las plantillas de chat estructuran las interacciones entre los usuarios y los modelos de IA, garantizando respuestas coherentes y apropiadas para el contexto. Incluyen componentes como prompts de sistema y mensajes basados en roles.</p> <h2 class="relative group"><a id="2-ajuste-fino-supervisado" 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="#2-ajuste-fino-supervisado"><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>2️⃣ Ajuste fino supervisado</span></h2> <p data-svelte-h="svelte-gxloiz">El ajuste fino supervisado (SFT) es un proceso crítico para adaptar modelos de lenguaje preentrenados a tareas específicas. Consiste en entrenar el modelo con un conjunto de datos específico de la tarea y ejemplos etiquetados. Para obtener una guía detallada sobre SFT, incluidos los pasos clave y las mejores prácticas, consulta <a href="https://huggingface.co/docs/trl/en/sft_trainer" rel="nofollow">la sección sobre ajuste fino supervisado de la documentación de TRL</a>.</p> <h2 class="relative group"><a id="3-adaptación-de-bajo-rango-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="#3-adaptación-de-bajo-rango-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>3️⃣ Adaptación de bajo rango (LoRA)</span></h2> <p data-svelte-h="svelte-1lqaa8o">La adaptación de bajo rango (LoRA) es una técnica para ajustar modelos de lenguaje añadiendo matrices de bajo rango a las capas del modelo. Esto permite un ajuste eficiente mientras se conserva el conocimiento preentrenado del modelo. Uno de los beneficios clave de LoRA es el importante ahorro de memoria que ofrece, lo que hace posible ajustar modelos grandes en hardware con recursos limitados.</p> <h2 class="relative group"><a id="4-evaluación" 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="#4-evaluación"><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>4️⃣ Evaluación</span></h2> <p data-svelte-h="svelte-1tt864b">La evaluación es un paso crucial en el proceso de ajuste. Nos permite medir el rendimiento del modelo en un conjunto de datos específico de la tarea.</p> <blockquote class="tip" data-svelte-h="svelte-kdzsys"><p>⚠️ Para aprovechar todas las funciones disponibles en el Hub de Modelos y 🤗 Transformers, recomendamos <a href="https://huggingface.co/join">crear una cuenta</a>.</p></blockquote> <h2 class="relative group"><a id="referencias" 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="#referencias"><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>Referencias</span></h2> <ul data-svelte-h="svelte-akupih"><li><a href="https://huggingface.co/docs/transformers/main/en/chat_templating" rel="nofollow">Documentación de Transformers sobre plantillas de chat</a></li> <li><a href="https://github.com/huggingface/trl/blob/main/trl/scripts/sft.py" rel="nofollow">Script para ajuste fino supervisado en TRL</a></li> <li><a href="https://huggingface.co/docs/trl/main/en/sft_trainer" rel="nofollow"><code>SFTTrainer</code> en TRL</a></li> <li><a href="https://arxiv.org/abs/2305.18290" rel="nofollow">Artículo sobre Direct Preference Optimization</a></li> <li><a href="https://huggingface.co/docs/trl/sft_trainer" rel="nofollow">Ajuste fino supervisado con TRL</a></li> <li><a href="https://github.com/huggingface/alignment-handbook" rel="nofollow">Cómo ajustar Google Gemma con ChatML y Hugging Face TRL</a></li> <li><a href="https://huggingface.co/learn/cookbook/en/fine_tuning_llm_to_generate_persian_product_catalogs_in_json_format" rel="nofollow">Ajuste de un LLM para generar catálogos de productos persas en formato JSON</a></li></ul> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/course/blob/main/chapters/es/chapter11/1.mdx" target="_blank"><svg class="mr-1" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M31,16l-7,7l-1.41-1.41L28.17,16l-5.58-5.59L24,9l7,7z"></path><path d="M1,16l7-7l1.41,1.41L3.83,16l5.58,5.59L8,23l-7-7z"></path><path d="M12.419,25.484L17.639,6.552l1.932,0.518L14.351,26.002z"></path></svg> <span data-svelte-h="svelte-zjs2n5"><span class="underline">Update</span> on GitHub</span></a> <p></p> | |
| <script> | |
| { | |
| __sveltekit_1nznq34 = { | |
| assets: "/docs/course/pr_1213/es", | |
| base: "/docs/course/pr_1213/es", | |
| env: {} | |
| }; | |
| const element = document.currentScript.parentElement; | |
| const data = [null,null]; | |
| Promise.all([ | |
| import("/docs/course/pr_1213/es/_app/immutable/entry/start.36d27295.js"), | |
| import("/docs/course/pr_1213/es/_app/immutable/entry/app.3b43d7f3.js") | |
| ]).then(([kit, app]) => { | |
| kit.start(app, element, { | |
| node_ids: [0, 21], | |
| data, | |
| form: null, | |
| error: null | |
| }); | |
| }); | |
| } | |
| </script> | |
Xet Storage Details
- Size:
- 16.5 kB
- Xet hash:
- fb55a1c723b3be62090df6e59baacae59b24daee110586775582b5c23f24dfc3
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.