|
|
--- |
|
|
language: |
|
|
- en |
|
|
- pl |
|
|
tags: |
|
|
- translation |
|
|
license: cc-by-4.0 |
|
|
datasets: |
|
|
- quickmt/quickmt-train.pl-en |
|
|
model-index: |
|
|
- name: quickmt-pl-en |
|
|
results: |
|
|
- task: |
|
|
name: Translation pol-eng |
|
|
type: translation |
|
|
args: pol-eng |
|
|
dataset: |
|
|
name: flores101-devtest |
|
|
type: flores_101 |
|
|
args: ell_Grek eng_Latn devtest |
|
|
metrics: |
|
|
- name: BLEU |
|
|
type: bleu |
|
|
value: 27.46 |
|
|
- name: CHRF |
|
|
type: chrf |
|
|
value: 57.18 |
|
|
- name: COMET |
|
|
type: comet |
|
|
value: 85.04 |
|
|
--- |
|
|
|
|
|
|
|
|
# `quickmt-pl-en` Neural Machine Translation Model |
|
|
|
|
|
`quickmt-pl-en` is a reasonably fast and reasonably accurate neural machine translation model for translation from `pl` into `en`. |
|
|
|
|
|
|
|
|
## 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`](https://github.com/eole-nlp/eole) |
|
|
* 195M parameter transformer 'big' with 8 encoder layers and 2 decoder layers |
|
|
* 20k separate Sentencepiece vocabs |
|
|
* Expested for fast inference to [CTranslate2](https://github.com/OpenNMT/CTranslate2) format |
|
|
* Training data: https://huggingface.co/datasets/quickmt/quickmt-train.pl-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: |
|
|
|
|
|
```bash |
|
|
git clone https://github.com/quickmt/quickmt.git |
|
|
pip install ./quickmt/ |
|
|
|
|
|
quickmt-model-download quickmt/quickmt-pl-en ./quickmt-pl-en |
|
|
``` |
|
|
|
|
|
Finally use the model in python: |
|
|
|
|
|
```python |
|
|
from quickmt import Translator |
|
|
|
|
|
# Auto-detects GPU, set to "cpu" to force CPU inference |
|
|
t = Translator("./quickmt-pl-en/", device="auto") |
|
|
|
|
|
# Translate - set beam size to 1 for faster speed (but lower quality) |
|
|
sample_text = 'Dr Ehud Ur, będący profesorem medycyny na Uniwersytecie Dalhousie w Halifaxie w Nowej Szkocji oraz przewodniczącym oddziału klinicznego i naukowego Kanadyjskiego Stowarzyszenia Cukrzycy, przestrzegł, iż badania nadal dopiero się zaczynają.' |
|
|
|
|
|
t(sample_text, beam_size=5) |
|
|
``` |
|
|
|
|
|
> 'Dr. Ehud Ur, a professor of medicine at Dalhousie University in Halifax, Nova Scotia and chairman of the clinical and scientific division of the Canadian Diabetes Association, warned that research is still just beginning.' |
|
|
|
|
|
```python |
|
|
# 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) |
|
|
``` |
|
|
|
|
|
> 'Professor of Medicine at Dalhous University Halifax in Nova Scotia, MD and Chair of the Canadian Diabetes Association’s Clinical and Scientific Division, cautioned that research is just beginning.' |
|
|
|
|
|
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](https://libretranslate.com/) 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](https://github.com/mjpost/sacrebleu) on the [Flores200 `devtest` test set](https://huggingface.co/datasets/facebook/flores) ("pol_Latn"->"eng_Latn"). `comet22` with the [`comet`](https://github.com/Unbabel/COMET) library and the [default model](https://huggingface.co/Unbabel/wmt22-comet-da). "Time (s)" is the time in seconds to translate the flores-devtest dataset (1012 sentences) on an RTX 4070s GPU with batch size 32 (faster speed is possible using a larger batch size). |
|
|
|
|
|
| | bleu | chrf2 | comet22 | Time (s) | |
|
|
|:---------------------------------|-------:|--------:|----------:|-----------:| |
|
|
| quickmt/quickmt-pl-en | 27.46 | 57.18 | 85.04 | 1.46 | |
|
|
| Helsinki-NLP/opus-mt-pl-en | 25.55 | 55.39 | 83.8 | 4.01 | |
|
|
| facebook/nllb-200-distilled-600M | 29.28 | 57.11 | 84.65 | 21.61 | |
|
|
| facebook/nllb-200-distilled-1.3B | 30.99 | 58.77 | 86.04 | 37.64 | |
|
|
| facebook/m2m100_418M | 22.12 | 52.51 | 80.41 | 17.99 | |
|
|
| facebook/m2m100_1.2B | 27.13 | 56.36 | 84.48 | 35.01 | |
|
|
|
