autotrain/pr_749/en/token_classification_params.html

<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Token Classification Parameters&quot;,&quot;local&quot;:&quot;token-classification-parameters&quot;,&quot;sections&quot;:[],&quot;depth&quot;:1}">
		<link href="/docs/autotrain/pr_749/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/entry/start.b4f8a0ef.js">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/chunks/scheduler.0219f8bd.js">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/chunks/singletons.74a96c49.js">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/chunks/paths.5815e531.js">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/entry/app.4f18d4a0.js">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/chunks/index.f61edf3b.js">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/nodes/0.3ba41ccf.js">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/nodes/32.f1df571d.js">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/chunks/CodeBlock.38e566ae.js">
		<link rel="modulepreload" href="/docs/autotrain/pr_749/en/_app/immutable/chunks/EditOnGithub.48fa589f.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Token Classification Parameters&quot;,&quot;local&quot;:&quot;token-classification-parameters&quot;,&quot;sections&quot;:[],&quot;depth&quot;:1}"><!-- HEAD_svelte-u9bgzb_END -->      <p></p>  <h1 class="relative group"><a id="token-classification-parameters" 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="#token-classification-parameters"><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>Token Classification Parameters</span></h1> <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">--batch-size BATCH_SIZE</span>
                    Training batch size <span class="hljs-keyword">to</span> use
<span class="hljs-comment">--seed SEED           Random seed for reproducibility</span>
<span class="hljs-comment">--epochs EPOCHS       Number of training epochs</span>
<span class="hljs-comment">--gradient_accumulation GRADIENT_ACCUMULATION</span>
                    Gradient accumulation steps
<span class="hljs-comment">--disable_gradient_checkpointing</span>
                    <span class="hljs-keyword">Disable</span> gradient checkpointing
<span class="hljs-comment">--lr LR               Learning rate</span>
<span class="hljs-comment">--log {none,wandb,tensorboard}</span>
                    Use experiment tracking
<span class="hljs-comment">--tokens-column TOKENS_COLUMN</span>
                    Tokens <span class="hljs-keyword">column</span> <span class="hljs-keyword">to</span> use. Must be a stringified list <span class="hljs-keyword">of</span> tokens <span class="hljs-keyword">if</span> <span class="hljs-keyword">using</span> a CSV file. <span class="hljs-keyword">Default</span> <span class="hljs-keyword">is</span> <span class="hljs-string">&#x27;tokens&#x27;</span>.
<span class="hljs-comment">--tags-column TAGS_COLUMN</span>
                    Tags <span class="hljs-keyword">column</span> <span class="hljs-keyword">to</span> use. Must be a stringified list <span class="hljs-keyword">of</span> tags <span class="hljs-keyword">if</span> <span class="hljs-keyword">using</span> a CSV file. <span class="hljs-keyword">Default</span> <span class="hljs-keyword">is</span> <span class="hljs-string">&#x27;tags&#x27;</span>.
<span class="hljs-comment">--max-seq-length MAX_SEQ_LENGTH</span>
                    <span class="hljs-keyword">Set</span> the maximum <span class="hljs-keyword">sequence</span> length (number <span class="hljs-keyword">of</span> tokens) that the model should handle <span class="hljs-keyword">in</span> a single <span class="hljs-keyword">input</span>. Longer <span class="hljs-keyword">sequences</span> are
                    truncated. Affects <span class="hljs-keyword">both</span> memory <span class="hljs-keyword">usage</span> <span class="hljs-keyword">and</span> computational requirements. <span class="hljs-keyword">Default</span> <span class="hljs-keyword">is</span> <span class="hljs-number">128</span> tokens.
<span class="hljs-comment">--warmup-ratio WARMUP_RATIO</span>
                    Define the proportion <span class="hljs-keyword">of</span> training <span class="hljs-keyword">to</span> be dedicated <span class="hljs-keyword">to</span> a linear warmup <span class="hljs-keyword">where</span> learning rate gradually increases. This can help
                    <span class="hljs-keyword">in</span> stabilizing the training process early <span class="hljs-keyword">on</span>. <span class="hljs-keyword">Default</span> ratio <span class="hljs-keyword">is</span> <span class="hljs-number">0.1</span>.
<span class="hljs-comment">--optimizer OPTIMIZER</span>
                    Choose the optimizer algorithm <span class="hljs-keyword">for</span> training the model. Different optimizers can affect the training speed <span class="hljs-keyword">and</span> model
                    performance. <span class="hljs-string">&#x27;adamw_torch&#x27;</span> <span class="hljs-keyword">is</span> used <span class="hljs-keyword">by</span> <span class="hljs-keyword">default</span>.
<span class="hljs-comment">--scheduler SCHEDULER</span>
                    <span class="hljs-keyword">Select</span> the learning rate scheduler <span class="hljs-keyword">to</span> adjust the learning rate based <span class="hljs-keyword">on</span> the number <span class="hljs-keyword">of</span> epochs. <span class="hljs-string">&#x27;linear&#x27;</span> decreases the
                    learning rate linearly <span class="hljs-keyword">from</span> the initial lr <span class="hljs-keyword">set</span>. <span class="hljs-keyword">Default</span> <span class="hljs-keyword">is</span> <span class="hljs-string">&#x27;linear&#x27;</span>. Try <span class="hljs-string">&#x27;cosine&#x27;</span> <span class="hljs-keyword">for</span> a cosine annealing schedule.
<span class="hljs-comment">--weight-decay WEIGHT_DECAY</span>
                    <span class="hljs-keyword">Set</span> the weight decay rate <span class="hljs-keyword">to</span> apply <span class="hljs-keyword">for</span> regularization. Helps <span class="hljs-keyword">in</span> preventing the model <span class="hljs-keyword">from</span> overfitting <span class="hljs-keyword">by</span> penalizing <span class="hljs-keyword">large</span>
                    weights. <span class="hljs-keyword">Default</span> <span class="hljs-keyword">is</span> <span class="hljs-number">0.0</span>, meaning <span class="hljs-keyword">no</span> weight decay <span class="hljs-keyword">is</span> applied.
<span class="hljs-comment">--max-grad-norm MAX_GRAD_NORM</span>
                    Specify the maximum norm <span class="hljs-keyword">of</span> the gradients <span class="hljs-keyword">for</span> gradient clipping. Gradient clipping <span class="hljs-keyword">is</span> used <span class="hljs-keyword">to</span> prevent the exploding gradient
                    problem <span class="hljs-keyword">in</span> deep neural networks. <span class="hljs-keyword">Default</span> <span class="hljs-keyword">is</span> <span class="hljs-number">1.0</span>.
<span class="hljs-comment">--logging-steps LOGGING_STEPS</span>
                    Determine how often <span class="hljs-keyword">to</span> <span class="hljs-keyword">log</span> training progress. <span class="hljs-keyword">Set</span> this <span class="hljs-keyword">to</span> the number <span class="hljs-keyword">of</span> steps <span class="hljs-keyword">between</span> <span class="hljs-keyword">each</span> <span class="hljs-keyword">log</span> output. <span class="hljs-number">-1</span> determines logging
                    steps automatically. <span class="hljs-keyword">Default</span> <span class="hljs-keyword">is</span> <span class="hljs-number">-1.</span>
<span class="hljs-comment">--eval-strategy {steps,epoch,no}</span>
                    Specify how often <span class="hljs-keyword">to</span> evaluate the model performance. <span class="hljs-keyword">Options</span> <span class="hljs-keyword">include</span> <span class="hljs-string">&#x27;no&#x27;</span>, <span class="hljs-string">&#x27;steps&#x27;</span>, <span class="hljs-string">&#x27;epoch&#x27;</span>. <span class="hljs-string">&#x27;epoch&#x27;</span> evaluates at the <span class="hljs-keyword">end</span> <span class="hljs-keyword">of</span>
                    <span class="hljs-keyword">each</span> training epoch <span class="hljs-keyword">by</span> <span class="hljs-keyword">default</span>.
<span class="hljs-comment">--save-total-limit SAVE_TOTAL_LIMIT</span>
                    <span class="hljs-keyword">Limit</span> the total number <span class="hljs-keyword">of</span> model checkpoints <span class="hljs-keyword">to</span> save. Helps manage disk space <span class="hljs-keyword">by</span> retaining <span class="hljs-keyword">only</span> the most recent checkpoints.
                    <span class="hljs-keyword">Default</span> <span class="hljs-keyword">is</span> <span class="hljs-keyword">to</span> save <span class="hljs-keyword">only</span> the latest one.
<span class="hljs-comment">--auto-find-batch-size</span>
                    <span class="hljs-keyword">Enable</span> automatic batch size determination based <span class="hljs-keyword">on</span> your hardware capabilities. <span class="hljs-keyword">When</span> <span class="hljs-keyword">set</span>, it tries <span class="hljs-keyword">to</span> find the largest batch
                    size that fits <span class="hljs-keyword">in</span> memory.
<span class="hljs-comment">--mixed-precision {fp16,bf16,None}</span>
                    Choose the <span class="hljs-type">precision</span> mode <span class="hljs-keyword">for</span> training <span class="hljs-keyword">to</span> optimize performance <span class="hljs-keyword">and</span> memory <span class="hljs-keyword">usage</span>. <span class="hljs-keyword">Options</span> are <span class="hljs-string">&#x27;fp16&#x27;</span>, <span class="hljs-string">&#x27;bf16&#x27;</span>, <span class="hljs-keyword">or</span> <span class="hljs-keyword">None</span> <span class="hljs-keyword">for</span>
                    <span class="hljs-keyword">default</span> <span class="hljs-type">precision</span>. <span class="hljs-keyword">Default</span> <span class="hljs-keyword">is</span> <span class="hljs-keyword">None</span>.<!-- HTML_TAG_END --></pre></div> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/autotrain-advanced/blob/main/docs/source/token_classification_params.mdx" 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_ewvlbq = {
						assets: "/docs/autotrain/pr_749/en",
						base: "/docs/autotrain/pr_749/en",
						env: {}
					};

					const element = document.currentScript.parentElement;

					const data = [null,null];

					Promise.all([
						import("/docs/autotrain/pr_749/en/_app/immutable/entry/start.b4f8a0ef.js"),
						import("/docs/autotrain/pr_749/en/_app/immutable/entry/app.4f18d4a0.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 32],
							data,
							form: null,
							error: null
						});
					});
				}
			</script>