michaelfeil / ct2fast-opus-mt-en-fr

huggingface.co
Total runs: 19
24-hour runs: 0
7-day runs: 0
30-day runs: 6
Model's Last Updated: May 19 2023
translation

Introduction of ct2fast-opus-mt-en-fr

Model Details of ct2fast-opus-mt-en-fr

# Fast-Inference with Ctranslate2

Speedup inference by 2x-8x using int8 inference in C++

quantized version of Helsinki-NLP/opus-mt-en-fr

pip install hf-hub-ctranslate2>=1.0.0 ctranslate2>=3.13.0

Converted using

ct2-transformers-converter --model Helsinki-NLP/opus-mt-en-fr --output_dir /home/michael/tmp-ct2fast-opus-mt-en-fr --force --copy_files README.md generation_config.json tokenizer_config.json vocab.json source.spm .gitattributes target.spm --quantization float16

Checkpoint compatible to ctranslate2 and hf-hub-ctranslate2

  • compute_type=int8_float16 for device="cuda"
  • compute_type=int8 for device="cpu"
from hf_hub_ctranslate2 import TranslatorCT2fromHfHub, GeneratorCT2fromHfHub
from transformers import AutoTokenizer

model_name = "michaelfeil/ct2fast-opus-mt-en-fr"
# use either TranslatorCT2fromHfHub or GeneratorCT2fromHfHub here, depending on model.
model = TranslatorCT2fromHfHub(
        # load in int8 on CUDA
        model_name_or_path=model_name, 
        device="cuda",
        compute_type="int8_float16",
        tokenizer=AutoTokenizer.from_pretrained("Helsinki-NLP/opus-mt-en-fr")
)
outputs = model.generate(
    text=["How do you call a fast Flan-ingo?", "User: How are you doing?"],
)
print(outputs)

Licence and other remarks:

This is just a quantized version. Licence conditions are intended to be idential to original huggingface repo.

Original description

opus-mt-en-fr
Benchmarks
testset BLEU chr-F
newsdiscussdev2015-enfr.en.fr 33.8 0.602
newsdiscusstest2015-enfr.en.fr 40.0 0.643
newssyscomb2009.en.fr 29.8 0.584
news-test2008.en.fr 27.5 0.554
newstest2009.en.fr 29.4 0.577
newstest2010.en.fr 32.7 0.596
newstest2011.en.fr 34.3 0.611
newstest2012.en.fr 31.8 0.592
newstest2013.en.fr 33.2 0.589
Tatoeba.en.fr 50.5 0.672

Runs of michaelfeil ct2fast-opus-mt-en-fr on huggingface.co

19
Total runs
0
24-hour runs
3
3-day runs
0
7-day runs
6
30-day runs

More Information About ct2fast-opus-mt-en-fr huggingface.co Model

More ct2fast-opus-mt-en-fr license Visit here:

https://choosealicense.com/licenses/apache-2.0

ct2fast-opus-mt-en-fr huggingface.co

ct2fast-opus-mt-en-fr huggingface.co is an AI model on huggingface.co that provides ct2fast-opus-mt-en-fr's model effect (), which can be used instantly with this michaelfeil ct2fast-opus-mt-en-fr model. huggingface.co supports a free trial of the ct2fast-opus-mt-en-fr model, and also provides paid use of the ct2fast-opus-mt-en-fr. Support call ct2fast-opus-mt-en-fr model through api, including Node.js, Python, http.

ct2fast-opus-mt-en-fr huggingface.co Url

https://huggingface.co/michaelfeil/ct2fast-opus-mt-en-fr

michaelfeil ct2fast-opus-mt-en-fr online free

ct2fast-opus-mt-en-fr huggingface.co is an online trial and call api platform, which integrates ct2fast-opus-mt-en-fr's modeling effects, including api services, and provides a free online trial of ct2fast-opus-mt-en-fr, you can try ct2fast-opus-mt-en-fr online for free by clicking the link below.

michaelfeil ct2fast-opus-mt-en-fr online free url in huggingface.co:

https://huggingface.co/michaelfeil/ct2fast-opus-mt-en-fr

ct2fast-opus-mt-en-fr install

ct2fast-opus-mt-en-fr is an open source model from GitHub that offers a free installation service, and any user can find ct2fast-opus-mt-en-fr on GitHub to install. At the same time, huggingface.co provides the effect of ct2fast-opus-mt-en-fr install, users can directly use ct2fast-opus-mt-en-fr installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

ct2fast-opus-mt-en-fr install url in huggingface.co:

https://huggingface.co/michaelfeil/ct2fast-opus-mt-en-fr

Url of ct2fast-opus-mt-en-fr

ct2fast-opus-mt-en-fr huggingface.co Url

Provider of ct2fast-opus-mt-en-fr huggingface.co

michaelfeil
ORGANIZATIONS

Other API from michaelfeil