File size: 617 Bytes
677d035
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
from datasets import load_dataset
from tokenizers import ByteLevelBPETokenizer

# load dataset
dataset = load_dataset("oscar", "unshuffled_deduplicated_id", split="train")

# Instantiate tokenizer
tokenizer = ByteLevelBPETokenizer()


def batch_iterator(batch_size=1000):
    for i in range(0, len(dataset), batch_size):
        yield dataset[i : i + batch_size]["text"]


# Customized training
tokenizer.train_from_iterator(
    batch_iterator(),
    vocab_size=50265,
    min_frequency=2,
    special_tokens=["<s>", "<pad>", "</s>", "<unk>", "<mask>",],
)

# Save files to disk
tokenizer.save(f"./tokenizer.json")