Commit
•
ae1d21d
1
Parent(s):
19adc42
Update README.md
Browse files
README.md
CHANGED
|
@@ -21,7 +21,6 @@ An extensive dataset with “artificial” errors was taken as a training corpus
|
|
| 21 |
- [SAGE library announcement](https://youtu.be/yFfkV0Qjuu0), DataFest 2023
|
| 22 |
- [Paper about synthetic error generation methods](https://www.dialog-21.ru/media/5914/martynovnplusetal056.pdf), Dialogue 2023
|
| 23 |
- [Paper about SAGE and our best solution](https://arxiv.org/abs/2308.09435), Review EACL 2024
|
| 24 |
-
- Path to model = "ai-forever/RuM2M100-1.2B"
|
| 25 |
|
| 26 |
|
| 27 |
### Examples
|
|
@@ -89,7 +88,7 @@ We compare our solution with both open automatic spell checkers and the ChatGPT
|
|
| 89 |
```python
|
| 90 |
from transformers import M2M100ForConditionalGeneration, M2M100Tokenizer
|
| 91 |
|
| 92 |
-
path_to_model = "
|
| 93 |
|
| 94 |
model = M2M100ForConditionalGeneration.from_pretrained(path_to_model)
|
| 95 |
tokenizer = M2M100Tokenizer.from_pretrained(path_to_model)
|
|
|
|
| 21 |
- [SAGE library announcement](https://youtu.be/yFfkV0Qjuu0), DataFest 2023
|
| 22 |
- [Paper about synthetic error generation methods](https://www.dialog-21.ru/media/5914/martynovnplusetal056.pdf), Dialogue 2023
|
| 23 |
- [Paper about SAGE and our best solution](https://arxiv.org/abs/2308.09435), Review EACL 2024
|
|
|
|
| 24 |
|
| 25 |
|
| 26 |
### Examples
|
|
|
|
| 88 |
```python
|
| 89 |
from transformers import M2M100ForConditionalGeneration, M2M100Tokenizer
|
| 90 |
|
| 91 |
+
path_to_model = "ai-forever/RuM2M100-1.2B"
|
| 92 |
|
| 93 |
model = M2M100ForConditionalGeneration.from_pretrained(path_to_model)
|
| 94 |
tokenizer = M2M100Tokenizer.from_pretrained(path_to_model)
|