Commit
•
bcfa1ab
1
Parent(s):
5465ae8
Update README.md
Browse files
README.md
CHANGED
@@ -1,12 +1,11 @@
|
|
1 |
---
|
2 |
license: mit
|
3 |
tags:
|
4 |
-
- generated_from_trainer
|
5 |
-
- natural-language-understanding
|
6 |
-
- nlu
|
7 |
- machine translation
|
8 |
- iva
|
9 |
- virtual assistants
|
|
|
|
|
10 |
metrics:
|
11 |
- bleu
|
12 |
model-index:
|
@@ -16,6 +15,7 @@ datasets:
|
|
16 |
- cartesinus/iva_mt_wslot
|
17 |
language:
|
18 |
- pl
|
|
|
19 |
---
|
20 |
|
21 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -47,8 +47,8 @@ BLEU was measured with [sacreBLEU](https://github.com/mjpost/sacrebleu) library.
|
|
47 |
|
48 |
## Model description, intended uses & limitations
|
49 |
|
50 |
-
Model is biased towards virtual assistant (IVA) sentences in prediction/translation. These sentences are short,
|
51 |
-
above results where WMT results are very low while in-domain test is very high.
|
52 |
|
53 |
This model will most probably force IVA translations on your text. As long as sentences that you are translating are more or less similar to massive and leyzer domains it
|
54 |
will be ok. If you will translate out-of-domain sentenences (such as for example News, Medical) that are not very similar then results will drop significantly up to the
|
|
|
1 |
---
|
2 |
license: mit
|
3 |
tags:
|
|
|
|
|
|
|
4 |
- machine translation
|
5 |
- iva
|
6 |
- virtual assistants
|
7 |
+
- natural-language-understanding
|
8 |
+
- nlu
|
9 |
metrics:
|
10 |
- bleu
|
11 |
model-index:
|
|
|
15 |
- cartesinus/iva_mt_wslot
|
16 |
language:
|
17 |
- pl
|
18 |
+
- en
|
19 |
---
|
20 |
|
21 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
47 |
|
48 |
## Model description, intended uses & limitations
|
49 |
|
50 |
+
Model is biased towards virtual assistant (IVA) sentences in prediction/translation. These sentences are short, imperatives with a lot of name entities (slots) and
|
51 |
+
particular vocabulary (for example settings name). It can be observed in above results where WMT results are very low while in-domain test is very high.
|
52 |
|
53 |
This model will most probably force IVA translations on your text. As long as sentences that you are translating are more or less similar to massive and leyzer domains it
|
54 |
will be ok. If you will translate out-of-domain sentenences (such as for example News, Medical) that are not very similar then results will drop significantly up to the
|