.ipynb_checkpoints/README-checkpoint.md

metadata

language:
  - en
  - cs
tags:
  - translation
license: cc-by-4.0
datasets:
  - quickmt/quickmt-train.cs-en
model-index:
  - name: quickmt-en-cs
    results:
      - task:
          name: Translation eng-ces
          type: translation
          args: eng-ces
        dataset:
          name: flores101-devtest
          type: flores_101
          args: eng_Latn ces_Latn devtest
        metrics:
          - name: BLEU
            type: bleu
            value: 33.73
          - name: CHRF
            type: chrf
            value: 60.29
          - name: COMET
            type: comet
            value: 88.77

quickmt-en-cs Neural Machine Translation Model

quickmt-en-cs is a reasonably fast and reasonably accurate neural machine translation model for translation from en into cs.

Try it on our Huggingface Space

Give it a try before downloading here: https://huggingface.co/spaces/quickmt/QuickMT-Demo

Model Information

• Trained using eole
• 200M parameter transformer 'big' with 8 encoder layers and 2 decoder layers
• 32k separate Sentencepiece vocabs
• Expested for fast inference to CTranslate2 format
• Training data: https://huggingface.co/datasets/quickmt/quickmt-train.lv-en/tree/main

See the eole model configuration in this repository for further details and the eole-model for the raw eole (pytorch) model.

Usage with quickmt

You must install the Nvidia cuda toolkit first, if you want to do GPU inference.

Next, install the quickmt python library and download the model:

git clone https://github.com/quickmt/quickmt.git
pip install ./quickmt/

quickmt-model-download quickmt/quickmt-en-cs ./quickmt-en-cs

Finally use the model in python:

from quickmt import Translator

# Auto-detects GPU, set to "cpu" to force CPU inference
t = Translator("./quickmt-en-cs/", device="auto")

# Translate - set beam size to 1 for faster speed (but lower quality)
sample_text = 'Dr. Ehud Ur, professor of medicine at Dalhousie University in Halifax, Nova Scotia and chair of the clinical and scientific division of the Canadian Diabetes Association cautioned that the research is still in its early days.'

t(sample_text, beam_size=5)

'Dr. Ehud Ur, profesor medicíny na Dalhousie University v Halifaxu v Novém Skotsku a předseda klinické a vědecké divize Canadian Diabetes Association varoval, že výzkum je stále v počátcích.'

# Get alternative translations by sampling
# You can pass any cTranslate2 `translate_batch` arguments
t([sample_text], sampling_temperature=1.2, beam_size=1, sampling_topk=50, sampling_topp=0.9)

'Profesor medicíny z Dalhousie University v Halifaxu v Novém Skotsku doktor Ehud Ur a předseda klinického a vědeckého oddělení Kanadské Diabetesové asociace upozornil, že jejich výzkum je stále ve svých počátcích.'

The model is in ctranslate2 format, and the tokenizers are sentencepiece, so you can use ctranslate2 directly instead of through quickmt. It is also possible to get this model to work with e.g. LibreTranslate which also uses ctranslate2 and sentencepiece. A model in safetensors format to be used with eole is also provided.

Metrics

bleu and chrf2 are calculated with sacrebleu on the Flores200 devtest test set ("eng_Latn"->"ces_Latn"). comet22 with the comet library and the default model. "Time (s)" is the time in seconds to translate the flores-devtest dataset (1012 sentences) on an Nvidia RTX 4070s GPU with batch size 32.

	bleu	chrf2	comet22	Time (s)
quickmt/quickmt-en-cs	33.73	60.29	88.77	1.35
Helsinki-NLP/opus-mt-en-cs	29.64	57.06	86.29	4.05
facebook/nllb-200-distilled-600M	28.39	56.53	88.74	25.7
facebook/nllb-200-distilled-1.3B	32.25	59.24	90.97	44.96
facebook/m2m100_418M	25.31	53.58	83.78	21.06
facebook/m2m100_1.2B	30.94	58.28	88.67	41.31