Update README.md
Browse files
README.md
CHANGED
@@ -13,7 +13,8 @@ pipeline_tag: translation
|
|
13 |
# Model Details
|
14 |
|
15 |
French-to-English Machine Translation model trained by Yasmin Moslem.
|
16 |
-
This
|
|
|
17 |
|
18 |
## Tools
|
19 |
|
@@ -25,6 +26,11 @@ This is a Transformer-based model originally trained with OpenNMT-py and then co
|
|
25 |
This model is trained on the French-to-English portion of the [UN Corpus](https://conferences.unite.un.org/UNCorpus/),
|
26 |
consisting of approx. 20 million segments.
|
27 |
|
|
|
|
|
|
|
|
|
|
|
28 |
## Demo
|
29 |
|
30 |
A demo of this model is available at: https://www.machinetranslation.io/
|
|
|
13 |
# Model Details
|
14 |
|
15 |
French-to-English Machine Translation model trained by Yasmin Moslem.
|
16 |
+
This model depends on the Transformer (base) architecture.
|
17 |
+
The model was originally trained with OpenNMT-py and then converted to the CTranslate2 format for efficient inference.
|
18 |
|
19 |
## Tools
|
20 |
|
|
|
26 |
This model is trained on the French-to-English portion of the [UN Corpus](https://conferences.unite.un.org/UNCorpus/),
|
27 |
consisting of approx. 20 million segments.
|
28 |
|
29 |
+
## Tokenizer
|
30 |
+
|
31 |
+
The tokenizer was trained using [SentencePiece](https://github.com/google/sentencepiece) on shared vocabulary.
|
32 |
+
Hence, there is only one SentencePiece model that can be used for tokenizing both the source and target texts.
|
33 |
+
|
34 |
## Demo
|
35 |
|
36 |
A demo of this model is available at: https://www.machinetranslation.io/
|