Update README.md
Browse files
README.md
CHANGED
|
@@ -5,13 +5,30 @@ language:
|
|
| 5 |
tags:
|
| 6 |
- babylm
|
| 7 |
- tokenizer
|
| 8 |
-
library_name: transformers
|
| 9 |
---
|
| 10 |
|
| 11 |
## Baby Tokenizer
|
| 12 |
|
| 13 |
Compact sentencepiece tokenizer for sample-efficient English language modeling, simply tokenizing natural language.
|
| 14 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
### Data
|
| 16 |
|
| 17 |
This tokeniser is derived from the BabyLM 100M dataset of mixed domain data, consisting of the following sources:
|
|
|
|
| 5 |
tags:
|
| 6 |
- babylm
|
| 7 |
- tokenizer
|
|
|
|
| 8 |
---
|
| 9 |
|
| 10 |
## Baby Tokenizer
|
| 11 |
|
| 12 |
Compact sentencepiece tokenizer for sample-efficient English language modeling, simply tokenizing natural language.
|
| 13 |
|
| 14 |
+
### Usage
|
| 15 |
+
|
| 16 |
+
#### Transformers
|
| 17 |
+
|
| 18 |
+
```py
|
| 19 |
+
from transformers import AutoTokenizer
|
| 20 |
+
|
| 21 |
+
tokenizer_baby = AutoTokenizer.from_pretrained("nilq/baby-tokenizer")
|
| 22 |
+
```
|
| 23 |
+
|
| 24 |
+
#### Tokenizers
|
| 25 |
+
|
| 26 |
+
```py
|
| 27 |
+
from tokenizers import Tokenizer
|
| 28 |
+
|
| 29 |
+
tokenizer_baby = Tokenizer.from_pretrained("nilq/baby-tokenizer")
|
| 30 |
+
```
|
| 31 |
+
|
| 32 |
### Data
|
| 33 |
|
| 34 |
This tokeniser is derived from the BabyLM 100M dataset of mixed domain data, consisting of the following sources:
|