trl/main/en/use_model.html

<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Use model after training&quot;,&quot;local&quot;:&quot;use-model-after-training&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Load and Generate&quot;,&quot;local&quot;:&quot;load-and-generate&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Use Adapters PEFT&quot;,&quot;local&quot;:&quot;use-adapters-peft&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}">
		<link href="/docs/trl/main/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/entry/start.183b226a.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/chunks/scheduler.85c25b89.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/chunks/singletons.98fe034d.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/chunks/paths.eb9df337.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/entry/app.9853b7f5.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/chunks/index.c142fe32.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/nodes/0.5efac18d.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/chunks/each.e59479a4.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/nodes/35.46d6ac36.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/chunks/CodeBlock.a5e95a57.js">
		<link rel="modulepreload" href="/docs/trl/main/en/_app/immutable/chunks/EditOnGithub.a592e7aa.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Use model after training&quot;,&quot;local&quot;:&quot;use-model-after-training&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Load and Generate&quot;,&quot;local&quot;:&quot;load-and-generate&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Use Adapters PEFT&quot;,&quot;local&quot;:&quot;use-adapters-peft&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"><!-- HEAD_svelte-u9bgzb_END -->      <p></p>  <h1 class="relative group"><a id="use-model-after-training" 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="#use-model-after-training"><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>Use model after training</span></h1> <p data-svelte-h="svelte-1f3c7jz">Once you have trained a model using either the SFTTrainer, PPOTrainer, or DPOTrainer, you will have a fine-tuned model that can be used for text generation. In this section, we’ll walk through the process of loading the fine-tuned model and generating text. If you need to run an inference server with the trained model, you can explore libraries such as <a href="https://github.com/huggingface/text-generation-inference" rel="nofollow"><code>text-generation-inference</code></a>.</p>  <h2 class="relative group"><a id="load-and-generate" 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-generate"><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 Generate</span></h2> <p data-svelte-h="svelte-jh3nqo">If you have fine-tuned a model fully, meaning without the use of PEFT you can simply load it like any other language model in transformers. E.g. the value head that was trained during the PPO training is no longer needed and if you load the model with the original transformer class it will be ignored:</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> AutoTokenizer, AutoModelForCausalLM

model_name_or_path = <span class="hljs-string">&quot;kashif/stack-llama-2&quot;</span> <span class="hljs-comment">#path/to/your/model/or/name/on/hub</span>
device = <span class="hljs-string">&quot;cpu&quot;</span> <span class="hljs-comment"># or &quot;cuda&quot; if you have a GPU</span>

model = AutoModelForCausalLM.from_pretrained(model_name_or_path).to(device)
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)

inputs = tokenizer.encode(<span class="hljs-string">&quot;This movie was really&quot;</span>, return_tensors=<span class="hljs-string">&quot;pt&quot;</span>).to(device)
outputs = model.generate(inputs)
<span class="hljs-built_in">print</span>(tokenizer.decode(outputs[<span class="hljs-number">0</span>]))<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-slosoc">Alternatively you can also use the pipeline:</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> pipeline

model_name_or_path = <span class="hljs-string">&quot;kashif/stack-llama-2&quot;</span> <span class="hljs-comment">#path/to/your/model/or/name/on/hub</span>
pipe = pipeline(<span class="hljs-string">&quot;text-generation&quot;</span>, model=model_name_or_path)
<span class="hljs-built_in">print</span>(pipe(<span class="hljs-string">&quot;This movie was really&quot;</span>)[<span class="hljs-number">0</span>][<span class="hljs-string">&quot;generated_text&quot;</span>])<!-- HTML_TAG_END --></pre></div>  <h2 class="relative group"><a id="use-adapters-peft" 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="#use-adapters-peft"><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>Use Adapters PEFT</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">from</span> peft <span class="hljs-keyword">import</span> PeftConfig, PeftModel
<span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoModelForCausalLM, AutoTokenizer

base_model_name = <span class="hljs-string">&quot;kashif/stack-llama-2&quot;</span> <span class="hljs-comment">#path/to/your/model/or/name/on/hub&quot;</span>
adapter_model_name = <span class="hljs-string">&quot;path/to/my/adapter&quot;</span>

model = AutoModelForCausalLM.from_pretrained(base_model_name)
model = PeftModel.from_pretrained(model, adapter_model_name)

tokenizer = AutoTokenizer.from_pretrained(base_model_name)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-11j5qxm">You can also merge the adapters into the base model so you can use the model like a normal transformers model, however the checkpoint will be significantly bigger:</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 = AutoModelForCausalLM.from_pretrained(base_model_name)
model = PeftModel.from_pretrained(model, adapter_model_name)

model = model.merge_and_unload()
model.save_pretrained(<span class="hljs-string">&quot;merged_adapters&quot;</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1hvquzo">Once you have the model loaded and either merged the adapters or keep them separately on top you can run generation as with a normal model outlined above.</p> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/trl/blob/main/docs/source/use_model.md" target="_blank"><span data-svelte-h="svelte-1kd6by1">&lt;</span> <span data-svelte-h="svelte-x0xyl0">&gt;</span> <span data-svelte-h="svelte-1dajgef"><span class="underline ml-1.5">Update</span> on GitHub</span></a>  <p></p> 
			
			<script>
				{
					__sveltekit_5yobsv = {
						assets: "/docs/trl/main/en",
						base: "/docs/trl/main/en",
						env: {}
					};

					const element = document.currentScript.parentElement;

					const data = [null,null];

					Promise.all([
						import("/docs/trl/main/en/_app/immutable/entry/start.183b226a.js"),
						import("/docs/trl/main/en/_app/immutable/entry/app.9853b7f5.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 35],
							data,
							form: null,
							error: null
						});
					});
				}
			</script>