jchevallard
commited on
Upload folder using huggingface_hub
Browse files- README.md +32 -0
- tokenizer.json +0 -0
- tokenizer_config.json +9 -0
README.md
ADDED
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: transformers
|
3 |
+
tags:
|
4 |
+
- transformers.js
|
5 |
+
- tokenizers
|
6 |
+
license: mit
|
7 |
+
---
|
8 |
+
|
9 |
+
Cloned from [Xenova/claude-tokenizer](https://huggingface.co/Xenova/claude-tokenizer)
|
10 |
+
|
11 |
+
|
12 |
+
# Claude Tokenizer
|
13 |
+
|
14 |
+
A 🤗-compatible version of the **Claude tokenizer** (adapted from [anthropics/anthropic-sdk-python](https://github.com/anthropics/anthropic-sdk-python)). This means it can be used with Hugging Face libraries including [Transformers](https://github.com/huggingface/transformers), [Tokenizers](https://github.com/huggingface/tokenizers), and [Transformers.js](https://github.com/xenova/transformers.js).
|
15 |
+
|
16 |
+
## Example usage:
|
17 |
+
|
18 |
+
### Transformers/Tokenizers
|
19 |
+
```py
|
20 |
+
from transformers import GPT2TokenizerFast
|
21 |
+
|
22 |
+
tokenizer = GPT2TokenizerFast.from_pretrained('Xenova/claude-tokenizer')
|
23 |
+
assert tokenizer.encode('hello world') == [9381, 2253]
|
24 |
+
```
|
25 |
+
|
26 |
+
### Transformers.js
|
27 |
+
```js
|
28 |
+
import { AutoTokenizer } from '@xenova/transformers';
|
29 |
+
|
30 |
+
const tokenizer = await AutoTokenizer.from_pretrained('Xenova/claude-tokenizer');
|
31 |
+
const tokens = tokenizer.encode('hello world'); // [9381, 2253]
|
32 |
+
```
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"add_prefix_space": false,
|
3 |
+
"bos_token": "<EOT>",
|
4 |
+
"clean_up_tokenization_spaces": true,
|
5 |
+
"eos_token": "<EOT>",
|
6 |
+
"model_max_length": 200000,
|
7 |
+
"tokenizer_class": "GPT2TokenizerFast",
|
8 |
+
"unk_token": "<EOT>"
|
9 |
+
}
|