Hugging Face's logo

Buckets:
hf-doc-build
/
doc-dev

Files
xet
hf-doc-build/doc-dev / diffusers /pr_14865 /en /training /jobs.html
HuggingFaceDocBuilder's picture
HuggingFaceDocBuilder
download
raw
32 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Hugging Face Jobs&quot;,&quot;local&quot;:&quot;hugging-face-jobs&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Train a LoRA on Jobs&quot;,&quot;local&quot;:&quot;train-a-lora-on-jobs&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Managing a Job&quot;,&quot;local&quot;:&quot;managing-a-job&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Train on your own images&quot;,&quot;local&quot;:&quot;train-on-your-own-images&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Train in FP8&quot;,&quot;local&quot;:&quot;train-in-fp8&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Run a script without a dependency header&quot;,&quot;local&quot;:&quot;run-a-script-without-a-dependency-header&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Save checkpoints for long runs&quot;,&quot;local&quot;:&quot;save-checkpoints-for-long-runs&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Next steps&quot;,&quot;local&quot;:&quot;next-steps&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/>
<link href="/docs/diffusers/main/en/_app/immutable/entry/start.CDWfyZ49.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/chunks/CLhuQgbj.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/chunks/wvPyzM3A.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/entry/app.SDxfIrXG.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/chunks/T1rQByR4.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/chunks/DnVIJfC5.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/chunks/DsnmJJEf.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/chunks/BeYo49jn.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/nodes/0.DK1AZNvC.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/nodes/2.DRQCmnKN.js" rel="modulepreload">
<link href="/docs/diffusers/main/en/_app/immutable/chunks/DCt3YxPl.js" rel="modulepreload">
<!--1ffuigs--><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Hugging Face Jobs&quot;,&quot;local&quot;:&quot;hugging-face-jobs&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Train a LoRA on Jobs&quot;,&quot;local&quot;:&quot;train-a-lora-on-jobs&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Managing a Job&quot;,&quot;local&quot;:&quot;managing-a-job&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Train on your own images&quot;,&quot;local&quot;:&quot;train-on-your-own-images&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Train in FP8&quot;,&quot;local&quot;:&quot;train-in-fp8&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Run a script without a dependency header&quot;,&quot;local&quot;:&quot;run-a-script-without-a-dependency-header&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Save checkpoints for long runs&quot;,&quot;local&quot;:&quot;save-checkpoints-for-long-runs&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Next steps&quot;,&quot;local&quot;:&quot;next-steps&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/><!---->
<link href="/docs/diffusers/main/en/_app/immutable/assets/0.tn0RQdqM.css" rel="modulepreload"> <!--[--><!--[0--><!--[--><!--[0--><!--[--><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> <!--[-1--><!--]--></div><!----> <!--[0--><h1 class="relative group"><a id="hugging-face-jobs" 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="#hugging-face-jobs"><span><svg 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>Hugging Face Jobs</span></h1><!--]--><!----> <p><a href="https://huggingface.co/docs/hub/jobs" rel="nofollow">Hugging Face Jobs</a> runs on Hugging Face GPUs, so you don’t need to set up a machine, and the run keeps going if you close your terminal. The Diffusers training scripts run on Jobs straight from their GitHub URL without needing to clone the repository or install anything locally.</p> <p>Before you start, follow the <a href="https://huggingface.co/docs/hub/jobs-quickstart" rel="nofollow">Jobs quickstart</a> to install the <code>hf</code> CLI, log in, and add credits to your account.</p> <!--[1--><h2 class="relative group"><a id="train-a-lora-on-jobs" 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-a-lora-on-jobs"><span><svg 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 a LoRA on Jobs</span></h2><!--]--><!----> <p>The command below trains a LoRA for <a href="https://huggingface.co/black-forest-labs/FLUX.2-klein-4B" rel="nofollow">FLUX.2-klein-4B</a> on the five dog photos in <a href="https://huggingface.co/datasets/diffusers/dog-example" rel="nofollow">diffusers/dog-example</a>. It runs on a single A10G and pushes the LoRA to your namespace on the Hub.</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 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="language-bash "><!----><span class="hljs-comment"># 20 steps is a test run. For a real run, raise --max_train_steps</span>
<span class="hljs-comment"># (the FLUX.2 README uses 500) and --timeout to match.</span>
hf <span class="hljs-built_in">jobs</span> uv run --flavor a10g-small --<span class="hljs-built_in">timeout</span> 30m -s HF_TOKEN -- \
https://raw.githubusercontent.com/huggingface/diffusers/main/examples/dreambooth/train_dreambooth_lora_flux2_klein.py \
--pretrained_model_name_or_path black-forest-labs/FLUX.2-klein-4B \
--dataset_name diffusers/dog-example \
--instance_prompt <span class="hljs-string">&quot;a photo of sks dog&quot;</span> \
--resolution 512 --mixed_precision bf16 --guidance_scale 1 \
--gradient_checkpointing --cache_latents \
--optimizer adamW --use_8bit_adam --learning_rate 1e-4 \
--max_train_steps 20 --seed 0 \
--output_dir /tmp/out \
--push_to_hub --hub_model_id your-username/klein-dog-lora<!----></pre></div><!----> <p><code>uv</code> installs the dependencies listed in the script’s <code># /// script</code> header, including Diffusers from <code>main</code>. Jobs don’t get a Hugging Face token by default, so <code>-s HF_TOKEN</code> forwards yours to let the script push the LoRA. The Job’s disk is discarded when the Job ends, and anything you don’t push is lost. The <a href="./dreambooth">DreamBooth</a> and <a href="./lora">LoRA</a> guides explain the training arguments.</p> <!--[1--><h2 class="relative group"><a id="managing-a-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="#managing-a-job"><span><svg 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>Managing a Job</span></h2><!--]--><!----> <p>The training logs print to your terminal while the Job runs. Closing the terminal or pressing <code>Ctrl+C</code> doesn’t stop training, because the Job runs on Hugging Face’s machines. To launch a long run without tying up your terminal, add <code>-d</code> before <code>--</code>. The command prints the Job ID and returns.</p> <p>Use the Job ID to reattach to the logs, check GPU memory and utilization while the model trains, or stop a run.</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 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="language-bash "><!---->hf <span class="hljs-built_in">jobs</span> logs -f &lt;job_id&gt;
hf <span class="hljs-built_in">jobs</span> stats &lt;job_id&gt;
hf <span class="hljs-built_in">jobs</span> cancel &lt;job_id&gt;<!----></pre></div><!----> <p>If the run fails, <code>hf jobs inspect &lt;job_id></code> shows the error message. See <a href="https://huggingface.co/docs/hub/jobs-manage" rel="nofollow">Manage Jobs</a> for the other commands.</p> <!--[1--><h2 class="relative group"><a id="train-on-your-own-images" 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-on-your-own-images"><span><svg 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 on your own images</span></h2><!--]--><!----> <p>To train on your images, mount their folder into the Job with <code>-v</code> and pass the mount path to <code>--instance_data_dir</code> instead of <code>--dataset_name</code>. Jobs uploads the folder to a private bucket before the Job starts and mounts it read-only. Uploading the same folder again only sends new or changed files.</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 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="language-bash "><!---->hf <span class="hljs-built_in">jobs</span> uv run --flavor a10g-small --<span class="hljs-built_in">timeout</span> 30m -s HF_TOKEN \
-v ./my-dog:/data -- \
https://raw.githubusercontent.com/huggingface/diffusers/main/examples/dreambooth/train_dreambooth_lora_flux2_klein.py \
--pretrained_model_name_or_path black-forest-labs/FLUX.2-klein-4B \
--instance_data_dir /data \
--instance_prompt <span class="hljs-string">&quot;a photo of sks dog&quot;</span> \
--resolution 512 --mixed_precision bf16 --guidance_scale 1 \
--gradient_checkpointing --cache_latents \
--optimizer adamW --use_8bit_adam --learning_rate 1e-4 \
--max_train_steps 20 --seed 0 \
--output_dir /tmp/out \
--push_to_hub --hub_model_id your-username/klein-dog-lora<!----></pre></div><!----> <p>The script opens every file in <code>--instance_data_dir</code> as an image, so keep only images in the folder. A hidden file such as <code>.DS_Store</code> makes the run fail. See <a href="https://huggingface.co/docs/hub/jobs-configuration#local-directories" rel="nofollow">Local directories</a> for more about mounting a folder.</p> <p>With <code>--with_prior_preservation</code>, the script generates class images into <code>--class_data_dir</code> before training starts. Point it at a path the Job can write to. To reuse the class images in later runs, put them in a read-write bucket mount like the one in <a href="#save-checkpoints-for-long-runs">Save checkpoints for long runs</a>.</p> <!--[1--><h2 class="relative group"><a id="train-in-fp8" 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-in-fp8"><span><svg 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 in FP8</span></h2><!--]--><!----> <p>The FLUX.2, Z-Image, Ideogram 4, and Krea 2 DreamBooth LoRA scripts take <code>--do_fp8_training</code> to train in FP8 with <a href="https://github.com/pytorch/ao" rel="nofollow">torchao</a>, which lowers memory use. No script header includes torchao, so add <code>--with torchao</code> before <code>--</code>. FP8 also needs a GPU with compute capability 8.9 or higher, such as an L4 (<code>l4x1</code>) or L40S (<code>l40sx1</code>). The A10G (8.6) and A100 (8.0) don’t support it.</p> <!--[1--><h2 class="relative group"><a id="run-a-script-without-a-dependency-header" 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="#run-a-script-without-a-dependency-header"><span><svg 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>Run a script without a dependency header</span></h2><!--]--><!----> <p>Only the scripts in <a href="https://github.com/huggingface/diffusers/tree/main/examples/dreambooth" rel="nofollow">examples/dreambooth</a> and <a href="https://github.com/huggingface/diffusers/tree/main/examples/advanced_diffusion_training" rel="nofollow">examples/advanced_diffusion_training</a> declare their dependencies in a <code># /// script</code> block at the top of the file. Some scripts in those folders, such as <code>train_dreambooth_lora_sdxl.py</code>, don’t have one.</p> <p>For a script without a header, pass each dependency with <code>--with</code>. Install Diffusers from source, because the training scripts require the development version. The other dependencies come from the script’s <code>requirements.txt</code> file and its imports.</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 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="language-bash "><!---->hf <span class="hljs-built_in">jobs</span> uv run --flavor a10g-small --<span class="hljs-built_in">timeout</span> 30m -s HF_TOKEN \
--with <span class="hljs-string">&quot;diffusers @ git+https://github.com/huggingface/diffusers.git&quot;</span> \
--with torch --with torchvision --with accelerate --with transformers \
--with peft --with datasets --with bitsandbytes \
--with ftfy --with tensorboard --with Jinja2 -- \
https://raw.githubusercontent.com/huggingface/diffusers/main/examples/dreambooth/train_dreambooth_lora_sdxl.py \
--pretrained_model_name_or_path stabilityai/stable-diffusion-xl-base-1.0 \
--pretrained_vae_model_name_or_path madebyollin/sdxl-vae-fp16-fix \
--dataset_name diffusers/dog-example \
--instance_prompt <span class="hljs-string">&quot;a photo of sks dog&quot;</span> \
--resolution 1024 --mixed_precision fp16 \
--gradient_checkpointing --use_8bit_adam --learning_rate 1e-4 \
--max_train_steps 20 --seed 0 \
--output_dir /tmp/out \
--push_to_hub --hub_model_id your-username/sdxl-dog-lora<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="save-checkpoints-for-long-runs" 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="#save-checkpoints-for-long-runs"><span><svg 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>Save checkpoints for long runs</span></h2><!--]--><!----> <p>A run that takes hours can time out or crash before it finishes. To keep its progress, mount a <a href="https://huggingface.co/docs/hub/storage-buckets" rel="nofollow">Storage Bucket</a> into the Job with <code>-v</code> and point <code>--output_dir</code> at it. Create the bucket first with <code>hf buckets create</code>.</p> <p>The script saves a <code>checkpoint-&lt;step></code> folder to <code>--output_dir</code> every <code>--checkpointing_steps</code> steps (500 by default), so the checkpoints land in the bucket as training goes. <code>--checkpoints_total_limit</code> caps how many are kept. <code>--resume_from_checkpoint latest</code> picks up from the newest checkpoint in <code>--output_dir</code>, and starts from scratch if there isn’t one, so the same command works for the first run and for each restart.</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 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="language-bash "><!---->hf <span class="hljs-built_in">jobs</span> uv run --flavor a10g-small --<span class="hljs-built_in">timeout</span> 8h -s HF_TOKEN \
-v hf://buckets/your-username/checkpoints:/ckpt -- \
https://raw.githubusercontent.com/huggingface/diffusers/main/examples/dreambooth/train_dreambooth_lora_flux2_klein.py \
--pretrained_model_name_or_path black-forest-labs/FLUX.2-klein-4B \
--dataset_name diffusers/dog-example \
--instance_prompt <span class="hljs-string">&quot;a photo of sks dog&quot;</span> \
--resolution 1024 --mixed_precision bf16 --guidance_scale 1 \
--gradient_checkpointing --cache_latents \
--optimizer adamW --use_8bit_adam --learning_rate 1e-4 \
--max_train_steps 5000 --seed 0 \
--checkpointing_steps 500 --checkpoints_total_limit 3 \
--resume_from_checkpoint latest \
--output_dir /ckpt/klein-dog<!----></pre></div><!----> <p>When training ends, the LoRA is saved to the bucket as <code>pytorch_lora_weights.safetensors</code>. See <a href="https://huggingface.co/docs/hub/jobs-configuration#volumes" rel="nofollow">Volumes</a> for the mount options.</p> <blockquote class="note"><p>Don’t add <code>--push_to_hub</code> when <code>--output_dir</code> holds checkpoints. The upload skips only <code>step_*</code> and <code>epoch_*</code> folders, so the <code>checkpoint-&lt;step></code> folders are pushed to the model repo along with the LoRA. Upload <code>pytorch_lora_weights.safetensors</code> to a model repo yourself instead.</p></blockquote> <!--[1--><h2 class="relative group"><a id="next-steps" 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="#next-steps"><span><svg 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>Next steps</span></h2><!--]--><!----> <ul><li>Read <a href="https://huggingface.co/docs/hub/jobs-training" rel="nofollow">Train Models on Jobs</a> in the Hub docs for the checks to run before a long job and how to read a failed one.</li> <li>Load your trained LoRA for inference with the <a href="../tutorials/using_peft_for_inference">LoRA</a> guide.</li> <li>Browse the <a href="https://github.com/huggingface/diffusers/tree/main/examples/dreambooth" rel="nofollow">DreamBooth README files</a> for model-specific commands and memory options.</li></ul> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/diffusers/blob/main/docs/source/en/training/jobs.md" target="_blank"><svg class="mr-1" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M31,16l-7,7l-1.41-1.41L28.17,16l-5.58-5.59L24,9l7,7z"></path><path d="M1,16l7-7l1.41,1.41L3.83,16l5.58,5.59L8,23l-7-7z"></path><path d="M12.419,25.484L17.639,6.552l1.932,0.518L14.351,26.002z"></path></svg><!----> <span><span class="underline">Update</span> on GitHub</span></a><!----> <p></p><!--]--><!----><!--]--><!--]--><!--]--> <!--[-1--><!--]--><!--]-->
<script>
{
__sveltekit_6eiw7u = {
base: "/docs/diffusers/main/en",
assets: "/docs/diffusers/main/en"
};
const element = document.currentScript.parentElement;
Promise.all([
import("/docs/diffusers/main/en/_app/immutable/entry/start.CDWfyZ49.js"),
import("/docs/diffusers/main/en/_app/immutable/entry/app.SDxfIrXG.js")
]).then(([kit, app]) => {
kit.start(app, element, {
node_ids: [0, 2],
data: [null,null],
form: null,
error: null
});
});
}
</script>

Xet Storage Details

Size:
32 kB
·
Xet hash:
55c345c2ccf00b29618cd731c752208613c45b52239e7fc80c7dec41235125cc

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