transformers.js/pr_1452/en/api/generation/parameters.md

# generation/parameters


* * *

<a id="module_generation/parameters..GenerationFunctionParameters" class="group"></a>

## `generation/parameters~GenerationFunctionParameters` : <code>Object</code>

**Kind**: inner typedef of [<code>generation/parameters</code>](#module_generation/parameters)  
<table>
  <thead>
    <tr>
      <th>Param</th><th>Type</th><th>Description</th>
    </tr>
  </thead>
  <tbody>
<tr>
    <td>[kwargs]</td><td><code>any</code></td><td><p>(<code>Dict[str, any]</code>, <em>optional</em>):</p>
</td>
    </tr>  </tbody>
</table>

**Properties**

<table>
  <thead>
    <tr>
      <th>Name</th><th>Type</th><th>Default</th><th>Description</th>
    </tr>
  </thead>
  <tbody>
<tr>
    <td>[inputs]</td><td><code>*</code></td><td><code></code></td><td><p>(<code>Tensor</code> of varying shape depending on the modality, <em>optional</em>):
The sequence used as a prompt for the generation or as model inputs to the encoder. If <code>null</code> the
method initializes it with <code>bos_token_id</code> and a batch size of 1. For decoder-only models <code>inputs</code>
should be in the format of <code>input_ids</code>. For encoder-decoder models <em>inputs</em> can represent any of
<code>input_ids</code>, <code>input_values</code>, <code>input_features</code>, or <code>pixel_values</code>.</p>
</td>
    </tr><tr>
    <td>[generation_config]</td><td><code>*</code></td><td><code></code></td><td><p>(<code>GenerationConfig</code>, <em>optional</em>):
The generation configuration to be used as base parametrization for the generation call.
<code>**kwargs</code> passed to generate matching the attributes of <code>generation_config</code> will override them.
If <code>generation_config</code> is not provided, the default will be used, which has the following loading
priority:</p>
<ul>
<li>(1) from the <code>generation_config.json</code> model file, if it exists;</li>
<li>(2) from the model configuration. Please note that unspecified parameters will inherit [<code>GenerationConfig</code>]&#39;s
default values, whose documentation should be checked to parameterize generation.</li>
</ul>
</td>
    </tr><tr>
    <td>[logits_processor]</td><td><code>*</code></td><td><code></code></td><td><p>(<code>LogitsProcessorList</code>, <em>optional</em>):
Custom logits processors that complement the default logits processors built from arguments and
generation config. If a logit processor is passed that is already created with the arguments or a
generation config an error is thrown. This feature is intended for advanced users.</p>
</td>
    </tr><tr>
    <td>[stopping_criteria]</td><td><code>*</code></td><td><code></code></td><td><p>(<code>StoppingCriteriaList</code>, <em>optional</em>):
Custom stopping criteria that complements the default stopping criteria built from arguments and a
generation config. If a stopping criteria is passed that is already created with the arguments or a
generation config an error is thrown. This feature is intended for advanced users.</p>
</td>
    </tr><tr>
    <td>[streamer]</td><td><code>*</code></td><td><code></code></td><td><p>(<code>BaseStreamer</code>, <em>optional</em>):
Streamer object that will be used to stream the generated sequences. Generated tokens are passed
through <code>streamer.put(token_ids)</code> and the streamer is responsible for any further processing.</p>
</td>
    </tr><tr>
    <td>[decoder_input_ids]</td><td><code>Array.&lt;number&gt;</code></td><td><code></code></td><td><p>(<code>number[]</code>, <em>optional</em>):
If the model is an encoder-decoder model, this argument is used to pass the <code>decoder_input_ids</code>.</p>
</td>
    </tr>  </tbody>
</table>


* * *



<EditOnGithub source="https://github.com/huggingface/transformers.js/blob/main/docs/source/api/generation/parameters.md" />