Hugging Face's logo

quickmt
/
quickmt-en-bn

Translation
English
Bengali
Eval Results
Model card Files Community
quickmt-en-bn
File size: 4,976 Bytes 
98c63bf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
 
---
language:
- en
- bn
tags:
- translation
license: cc-by-4.0
datasets:
- quickmt/quickmt-train.bn-en
model-index:
- name: quickmt-en-bn
  results:
  - task:
      name: Translation eng-ben
      type: translation
      args: eng-ben
    dataset:
      name: flores101-devtest
      type: flores_101
      args: eng_Latn ben_Beng devtest
    metrics:
    - name: BLEU
      type: bleu
      value: 19.27
    - name: CHRF
      type: chrf
      value: 53.34
    - name: COMET
      type: comet
      value: 84.68
---


# `quickmt-en-bn` Neural Machine Translation Model 

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


## Model Information

* Trained using [`eole`](https://github.com/eole-nlp/eole)
* 185M parameter transformer 'big' with 8 encoder layers and 2 decoder layers
* 20k sentencepiece vocabularies
* Exported for fast inference to [CTranslate2](https://github.com/OpenNMT/CTranslate2) format
* Training data: https://huggingface.co/datasets/quickmt/quickmt-train.bn-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-en-bn ./quickmt-en-bn
```

Finally use the model in python:

```python
from quickmt import Translator

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

# Translate - set beam size to 5 for higher quality (but slower speed)
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)
```

> 'নোভা স্কটিয়ার হ্যালিফ্যাক্সের ডালহৌসি বিশ্ববিদ্যালয়ের মেডিসিনের অধ্যাপক এবং কানাডিয়ান ডায়াবেটিস অ্যাসোসিয়েশনের ক্লিনিকাল ও বৈজ্ঞানিক বিভাগের চেয়ারম্যান ড. এহুদ উর সতর্ক করেছেন যে গবেষণাটি এখনও তার প্রাথমিক দিনগুলিতে রয়েছে।'

```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)
```

> 'নোভা স্কোশিয়ার হ্যালিফ্যাক্সে ডালহৌসি বিশ্ববিদ্যালয়ের মেডিসিনের প্রফেসর ও কানাডিয়ান ডায়াবেটিস এসোসিয়েশনের ক্লিনিক্যাল অ্যান্ড সায়েন্টিফিক বিভাগের সভাপতি ড. এহুদ উর সাবধান করেছেন যে গবেষণা এখনও শুরুর দিকেই রয়েছে।'


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`.


## 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) ("eng_Latn"->"ben_Beng"). `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 (using `ctranslate2`) the flores-devtest dataset (1012 sentences) on an RTX 4070s GPU with batch size 32 (faster speed is possible with a larger batch size).

|                                      |   bleu |   chrf2 |   comet22 |   Time (s) |
|:-------------------------------------|-------:|--------:|----------:|-----------:|
| quickmt/quickmt-en-bn                |  19.27 |   53.34 |     84.68 |       1.17 |
| facebook/nllb-200-distilled-600M     |  15.64 |   50.65 |     85.69 |      28.01 |
| facebook/nllb-200-distilled-1.3B     |  17.31 |   52.17 |     86.45 |      46.94 |
| facebook/m2m100_418M                 |   7.39 |   36.54 |     70.44 |      44.3  |
| facebook/m2m100_1.2B                 |  12    |   45.19 |     81.92 |      61.97 |