wavey-ai / mel-spec Goto Github PK

View Code? Open in Web Editor NEW

36.0 36.0 9.0 34.1 MB

Rust library for whisper.cpp compatible Mel spectrograms

License: MIT License

Rust 100.00%

mel-spec's People

Contributors

Stargazers

Watchers

Forkers

xd009642 fl33tw00d katya4oyu skewballfox dictate-your-destiny akrichikov wfjsw ezzatomar mrienstra

mel-spec's Issues

Expose a library API

So I was looking at using this as a library but none of the functionality seems publicly exposed? The readme looks quite impressive compared to the API functions docs.rs shows for the crate https://docs.rs/mel_spec/latest/mel_spec/index.html

Error running the browser example

Hi, I'm not really familiar with rust.

When running the browser example I got this error once transcription starts.

Looking in mel_spec_pipeline.js I find this.

I on rustc version 1.78.0, cargo 1.78.0. MacBook M1 Pro.

When building I had some warnings but nothing dramatic.
Not sure where to start.

Am I doing something wrong? Any help is welcomed. :)

Help with pcm to mel_spectrograms in the non buffered/streamed case

I am trying to use the library to convert 4 second chunks of pcm data (at 16kHz) to mel spectrograms.
The library has some really nice tools for streaming data, what would you recommend when I have the exact number of samples in a Vec?

More simply I am looking for something that looks like this:

let pcm = vec![1.0; 4 * 16000];
let mel = pcm_to_mel(pcm, n_mel, fft_size, hop_size);

I am reasonably new to audio/rust in general and have been struggling to find a library with this functionality.
Everything seems to be fancier/more complex than I need.

Thanks,
Sefton

interested in testing against data generated by librosa's functions

Awesome project.
If I was able to generate data from librosa's python functions(input and output npy files) to compare against, would you be interested in including those as a test case?

I started working on a tool for generating files for this for the library I've been working on after realizing none of the existing python functions for computing an MFCC had anywhere near the same results, and models may be somewhat overfit to the preprocessing step. but I'm less of a mathematician and more of a code monkey, so I'm just guessing.

Also interested in contributors?

mel_spec::mel returns NaNs

#[test]
fn test_mel() {
    let mel = mel(16000.0f64, 1024, 128, true, true);

    println!("{:?}", mel);
}

---- tests::pitch::tests::test_mel stdout ----
[[NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN],
 [NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN],
 [NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN],
 [NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN],
 [NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN],
 ...,
 [NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN],
 [NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN],
 [NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN],
 [NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN],
 [NaN, NaN, NaN, NaN, NaN, ..., NaN, NaN, NaN, NaN, NaN]], shape=[128, 513], strides=[513, 1], layout=Cc (0x5), const ndim=2

wavey-ai / mel-spec Goto Github PK

mel-spec's People

Contributors

Stargazers

Watchers

Forkers

mel-spec's Issues

Expose a library API

Error running the browser example

Help with pcm to mel_spectrograms in the non buffered/streamed case

interested in testing against data generated by librosa's functions

mel_spec::mel returns NaNs

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent

Jobs