gogs
/
rust-whisper
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

no more rwhisper

master v0.2.4
Bel LaPointe 2023-12-21 15:07:05 -05:00
parent 3b79024bdd
commit f326f077ce
3 changed files with 15 additions and 2301 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2248
rust-whisper-lib/Cargo.lock generated
View File

File diff suppressed because it is too large Load Diff

3
rust-whisper-lib/Cargo.toml
View File

@ -12,6 +12,3 @@ tokio = "1.27"
byteorder = "1.5.0"
chrono = "0.4.31"
clap = { version = "4.4.10", features = ["derive"] }
rwhisper = "0.1.2"
rodio = "0.17.3"
futures = "0.3.29"

65
rust-whisper-lib/src/lib.rs
View File

@ -3,7 +3,6 @@ use clap::Parser;
use std::thread;
use std::fs::File;
use std::io::Write;
use rwhisper;
#[derive(Parser, Debug, Clone)]
pub struct Flags {
@ -258,54 +257,6 @@ impl Engine {
}
}
struct Engine2 {
model: rwhisper::Whisper,
}
fn new_engine2(model_path: Option<String>, model_buffer: Option<Vec<u8>>, threads: i32) -> Result<Engine2, String> {
match rwhisper::WhisperBuilder::default()
.with_cpu(true)
.with_language(Some(rwhisper::WhisperLanguage::English))
.with_source(rwhisper::WhisperSource::TinyEn)
.build() {
Ok(model) => Ok(Engine2{model: model}),
Err(msg) => Err(format!("failed to create model: {}", msg)),
}
}
impl Engine2 {
fn transcribe(&self, data: &Vec<f32>) -> Result<Transcribed, String> {
self._transcribe(data)
}
fn _transcribe(&self, data: &Vec<f32>) -> Result<Transcribed, String> {
let buffer = rodio::buffer::SamplesBuffer::new(1, 16_000, data.clone());
let stream = self.model.transcribe(buffer);
if stream.as_ref().is_err() {
return Err(format!("failed to start transcribing: {}", stream.err().unwrap()));
}
let stream = stream.unwrap();
let future = async {
let mut w: Vec<u8> = vec![];
stream.write_to(&mut w).await.unwrap();
w
};
let w = futures::executor::block_on(future);
let mut result = new_whispered();
result.push(String::from_utf8(w).unwrap(), 0, (100 * data.len() / 16_000) as i64);
Ok(result)
/*
let mut w: Vec<u8> = vec![];
stream.write_to(&mut w).await.unwrap();
let mut result = new_whispered();
result.push(String::from_utf8(w).unwrap(), 0, (100 * data.len() / 16_000) as i64);
Ok(result)
*/
}
}
struct ATranscribe {
data: Vec<f32>,
ack: Option<std::sync::mpsc::SyncSender<bool>>,
@ -404,20 +355,4 @@ mod tests {
"../gitea-whisper-rs/sys/whisper.cpp/bindings/go/samples/jfk.wav".to_string(),
);
}
#[test]
fn test_transcribe_tiny_jfk_wav_rwhisper() {
let engine_2 = new_engine2(
Some("../models/ggml-tiny.en.bin".to_string()),
None,
4,
).expect("failed to make new engine2");
let data = f32_from_wav_file(&"../gitea-whisper-rs/sys/whisper.cpp/bindings/go/samples/jfk.wav".to_string()).expect("failed to read jfk.wav");
let start = std::time::Instant::now();
for i in 0..2 {
let result = engine_2.transcribe(&data).expect("failed to transcribe");
println!("rwhisper = {}s", start.elapsed().as_secs_f32());
assert_eq!(" And so my fellow American asked not what your country can do for you, ask what you can do for your country.".to_string(), result.to_string());
}
}
}