ctranslate2-rs

This library provides Rust bindings for OpenNMT/CTranslate2. At this time, it has only been tested and confirmed to work on macOS and Linux. Windows support is available experimentally, but it has not been thoroughly tested and may have limitations or require additional configuration.

Compilation

If you plan to use GPU acceleration, CUDA and cuDNN are available. Please enable the cuda or cudnn feature and set the CUDA_TOOLKIT_ROOT_DIR environment variable appropriately.

Several backends are available for use: OpenBLAS, Intel MKL, Ruy, and Apple Accelerate.

• OpenBLAS: To use OpenBLAS, enable the openblas feature and add the path to the directory containing libopenblas.a to the LIBRARY_PATH environment variable.
• Intel MKL: To use Intel MKL, enable the mkl feature and set the path to the Intel libraries in the MKLROOT environment variable (default is /opt/intel).
• Ruy: To use Ruy, enable the ruy feature.
• Apple Accelerate: Available only on macOS, enable the accelerate feature to use Apple Accelerate.

If no feature is specified:

• On macOS, Apple Accelerate will be used by default.
• On Linux and Windows, Ruy will be used by default.

The installation of CMake is required to compile the library.

Additional notes for Windows: it is necessary to add RUSTFLAGS=-C target-feature=+crt-static to the environment variables for compilation.

Model Conversion for CTranslate2

To use model files with CTranslate2, they must first be converted. Below is an example of how to convert the nllb-200-distilled-600M model:

pip install ctranslate2
ct2-transformers-converter --model facebook/nllb-200-distilled-600M --output_dir nllb-200-distilled-600M --copy_files tokenizer.json

For more details, please refer to the CTranslate2's docs.

Example of text translation

The following example translates English to German and Japanese using the previously converted model nllb-200-distilled-600M.

use anyhow::Result;

use ct2rs::config::{Config, Device};
use ct2rs::{TranslationOptions, Translator};
use ct2rs::tokenizers::Tokenizer;

fn main() -> Result<()> {
    let path = "/path/to/nllb-200-distilled-600M";
    let t = Translator::new(&path, Tokenizer::new(&path)?, &Config::default())?;
    let res = t.translate_batch_with_target_prefix(
        &vec![
            "Hello world!",
            "This library provides Rust bindings for CTranslate2.",
        ],
        &vec![vec!["deu_Latn"], vec!["jpn_Jpan"]],
        &TranslationOptions {
            return_scores: true,
            ..Default::default()
        },
    )?;
    for r in res {
        println!("{}, (score: {:?})", r.0, r.1);
    }


    Ok(())
}

Output

Hallo Welt!<unk>, (score: Some(-0.5597002))
このライブラリでは,CTranslate2 の Rust バインディングが提供されています., (score: Some(-0.56321025))

Example of text generation

use anyhow::Result;

use ct2rs::config::{Config, Device};
use ct2rs::{Generator, GenerationOptions};
use ct2rs::sentencepiece::Tokenizer;

fn main() -> Result<()> {
    let path = "/path/to/model";
    let g = Generator::new(&path, Tokenizer::new(&path)?, &Config::default())?;
    let res = g.generate_batch(
        &vec!["prompt"],
        &GenerationOptions::default(),
    )?;
    for r in res {
        println!("{:?}", r.0);
    }

    Ok(())
}

License

This application is released under the MIT License. For details, see the LICENSE file.