add text chunking for long-form generation (Fixes #4)
This commit is contained in:
@@ -186,15 +186,15 @@ async function generateSpeech() {
|
||||
`;
|
||||
|
||||
const totalStep = parseInt(totalStepInput.value);
|
||||
const textList = [text];
|
||||
|
||||
showStatus('ℹ️ <strong>Generating speech from text...</strong>');
|
||||
const tic = Date.now();
|
||||
|
||||
const { wav, duration } = await textToSpeech.call(
|
||||
textList,
|
||||
text,
|
||||
currentStyle,
|
||||
totalStep,
|
||||
totalStep,
|
||||
0.3,
|
||||
(step, total) => {
|
||||
showStatus(`ℹ️ <strong>Denoising (${step}/${total})...</strong>`);
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user