Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
- et
|
| 5 |
+
tags:
|
| 6 |
+
- translation
|
| 7 |
+
license: cc-by-4.0
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
### Translation model for en-et OPUS_HPLT v1.0
|
| 11 |
+
|
| 12 |
+
This repository contains the model weights for translation models trained with Marian for HPLT project. For usage instructions, evaluation scripts, and inference scripts, please refer to the [HPLT-MT-Models v1.0](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0) GitHub repository.
|
| 13 |
+
|
| 14 |
+
* Source language: en
|
| 15 |
+
* Target language: et
|
| 16 |
+
* Dataset: All of OPUS including HPLT
|
| 17 |
+
* Model: transformer-base
|
| 18 |
+
* Tokenizer: SentencePiece (Unigram)
|
| 19 |
+
* Cleaning: We use OpusCleaner for cleaning the corpus. Details about rules used can be found in the filter files in [Github](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0/data/en-et/raw/v2)
|
| 20 |
+
|
| 21 |
+
To run inference with Marian, refer to the [Inference/Decoding/Translation](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0#inferencedecodingtranslation) section of our GitHub repository.
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
## Benchmarks
|
| 25 |
+
|
| 26 |
+
| testset | BLEU | chr-F | comet |
|
| 27 |
+
| -------------------------------------- | ---- | ----- | ----- |
|
| 28 |
+
| flores200 | 25.2 | 54.8 | 0.8994 |
|
| 29 |
+
| ntrex | 24.5 | 53.1 | 0.8687 |
|