File size: 899 Bytes
8b9ba87
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
28
29
30
31
32
#!/usr/bin/env python3
from datasets import load_dataset
from tokenizers import ByteLevelBPETokenizer

# Load dataset
<<<<<<< HEAD
dataset = load_dataset("oscar", "unshuffled_deduplicated_es", split="train[:5000000]")

# Instantiate tokenizer
tokenizer = ByteLevelBPETokenizer()
def batch_iterator(batch_size=100_000):
=======
dataset = load_dataset("oscar", "unshuffled_deduplicated_es", split="train")

# Instantiate tokenizer
tokenizer = ByteLevelBPETokenizer()
def batch_iterator(batch_size=1_000_000):
>>>>>>> d5cede47e74aa6ec36f20acf5aba37c6734c6186
    for i in range(0, len(dataset), batch_size):
        yield dataset["text"][i: i + batch_size]

# 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("./tokenizer.json")