huggingface/candle
1
0
Fork 0
mirror of https://github.com/huggingface/candle.git synced 2025-06-15 18:28:24 +00:00
Code Issues Packages Projects Releases Wiki Activity

Preliminary support for whisper v3. (#1294)

Browse Source 
* Preliminary support for whisper v3.

* Add the missing files.
This commit is contained in:
Laurent Mazare
2023-11-08 06:42:52 +01:00
committed by GitHub
parent f3a4f3db76
commit 2d28497197
5 changed files with 31 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
candle-wasm-examples/whisper/src/audio.rs
View File

@ -200,6 +200,7 @@ fn log_mel_spectrogram_<T: Float + std::fmt::Display>(
}
pub fn pcm_to_mel<T: Float + std::fmt::Display>(
cfg: &worker::m::Config,
samples: &[T],
filters: &[T],
) -> anyhow::Result<Vec<T>> {
@ -208,7 +209,7 @@ pub fn pcm_to_mel<T: Float + std::fmt::Display>(
filters,
worker::m::N_FFT,
worker::m::HOP_LENGTH,
worker::m::N_MELS,
cfg.num_mel_bins,
false,
);
Ok(mel)

5
candle-wasm-examples/whisper/src/worker.rs
View File

@ -349,9 +349,10 @@ impl Decoder {
.map(|v| *v as f32 / 32768.)
.collect();
console_log!("pcm data loaded {}", pcm_data.len());
let mel = crate::audio::pcm_to_mel(&pcm_data, &self.mel_filters)?;
let mel = crate::audio::pcm_to_mel(self.model.config(), &pcm_data, &self.mel_filters)?;
let mel_len = mel.len();
let mel = Tensor::from_vec(mel, (1, m::N_MELS, mel_len / m::N_MELS), &device)?;
let n_mels = self.model.config().num_mel_bins;
let mel = Tensor::from_vec(mel, (1, n_mels, mel_len / n_mels), &device)?;
console_log!("loaded mel: {:?}", mel.dims());
let segments = self.run(&mel)?;
Ok(segments)