Create README
Browse files
README.md
CHANGED
@@ -1 +1,16 @@
|
|
1 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
def DummyTok(model_max_length=4):
|
2 |
+
import tempfile
|
3 |
+
|
4 |
+
from tokenizers import Tokenizer, models
|
5 |
+
from transformers.tokenization_utils_fast import PreTrainedTokenizerFast
|
6 |
+
|
7 |
+
vocab = [(chr(i), i) for i in range(256)]
|
8 |
+
tokenizer = Tokenizer(models.Unigram(vocab))
|
9 |
+
with tempfile.NamedTemporaryFile() as f:
|
10 |
+
tokenizer.save(f.name)
|
11 |
+
real_tokenizer = PreTrainedTokenizerFast(
|
12 |
+
tokenizer_file=f.name, model_max_length=model_max_length
|
13 |
+
)
|
14 |
+
|
15 |
+
real_tokenizer.save("dummy_tokenizer.json")
|
16 |
+
return real_tokenizer
|