From ff3019fc052a8df575b3e30ff4b0e8e85a76255c Mon Sep 17 00:00:00 2001
From: Joshua Lochner <admin@xenova.com>
Date: Wed, 6 Dec 2023 22:17:11 +0200
Subject: [PATCH] Add example usage for `SpeechT5ForSpeechToText` (#438)

---
 src/models.js | 39 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 39 insertions(+)

diff --git a/src/models.js b/src/models.js
index 1c55d88..48fab2f 100644
--- a/src/models.js
+++ b/src/models.js
@@ -3859,6 +3859,43 @@ export class SpeechT5Model extends SpeechT5PreTrainedModel { };
 
 /**
  * SpeechT5 Model with a speech encoder and a text decoder.
+ * 
+ * **Example:** Generate speech from text with `SpeechT5ForSpeechToText`.
+ * ```javascript
+ * import { AutoTokenizer, AutoProcessor, SpeechT5ForTextToSpeech, SpeechT5HifiGan, Tensor } from '@xenova/transformers';
+ * 
+ * // Load the tokenizer and processor
+ * const tokenizer = await AutoTokenizer.from_pretrained('Xenova/speecht5_tts');
+ * const processor = await AutoProcessor.from_pretrained('Xenova/speecht5_tts');
+ * 
+ * // Load the models
+ * // NOTE: We use the unquantized versions as they are more accurate
+ * const model = await SpeechT5ForTextToSpeech.from_pretrained('Xenova/speecht5_tts', { quantized: false });
+ * const vocoder = await SpeechT5HifiGan.from_pretrained('Xenova/speecht5_hifigan', { quantized: false });
+ * 
+ * // Load speaker embeddings from URL
+ * const speaker_embeddings_data = new Float32Array(
+ *     await (await fetch('https://huggingface.co/datasets/Xenova/transformers.js-docs/resolve/main/speaker_embeddings.bin')).arrayBuffer()
+ * );
+ * const speaker_embeddings = new Tensor(
+ *     'float32',
+ *     speaker_embeddings_data,
+ *     [1, speaker_embeddings_data.length]
+ * )
+ * 
+ * // Run tokenization
+ * const { input_ids } = tokenizer('Hello, my dog is cute');
+ * 
+ * // Generate waveform
+ * const { waveform } = await model.generate_speech(input_ids, speaker_embeddings, { vocoder });
+ * console.log(waveform)
+ * // Tensor {
+ * //   dims: [ 26112 ],
+ * //   type: 'float32',
+ * //   size: 26112,
+ * //   data: Float32Array(26112) [ -0.00043630177970044315, -0.00018082228780258447, ... ],
+ * // }
+ * ```
  */
 export class SpeechT5ForSpeechToText extends SpeechT5PreTrainedModel { }
 
@@ -3984,6 +4021,8 @@ export class SpeechT5ForTextToSpeech extends SpeechT5PreTrainedModel {
 
 /**
  * HiFi-GAN vocoder.
+ * 
+ * See [SpeechT5ForSpeechToText](./models#module_models.SpeechT5ForSpeechToText) for example usage.
  */
 export class SpeechT5HifiGan extends PreTrainedModel {
     main_input_name = 'spectrogram';