mariagrandury
commited on
Commit
•
8bb6457
1
Parent(s):
bce6976
add tokenizer
Browse files- special_tokens_map.json +6 -0
- tokenizer_config.json +9 -0
- vocab.json +33 -0
special_tokens_map.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"eos_token": "</s>",
|
4 |
+
"unk_token": "<unk>",
|
5 |
+
"pad_token": "<pad>"
|
6 |
+
}
|
tokenizer_config.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"unk_token": "<unk>",
|
3 |
+
"bos_token": "<s>",
|
4 |
+
"eos_token": "</s>",
|
5 |
+
"pad_token": "<pad>",
|
6 |
+
"do_lower_case": false,
|
7 |
+
"return_attention_mask": false,
|
8 |
+
"do_normalize": true
|
9 |
+
}
|
vocab.json
ADDED
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"<pad>": 0,
|
3 |
+
"<s>": 1,
|
4 |
+
"</s>": 2,
|
5 |
+
"<unk>": 3,
|
6 |
+
"E": 4,
|
7 |
+
"A": 5,
|
8 |
+
"O": 6,
|
9 |
+
"S": 7,
|
10 |
+
"R": 8,
|
11 |
+
"N": 9,
|
12 |
+
"I": 10,
|
13 |
+
"D": 11,
|
14 |
+
"L": 12,
|
15 |
+
"C": 13,
|
16 |
+
"T": 14,
|
17 |
+
"U": 15,
|
18 |
+
"M": 16,
|
19 |
+
"P": 17,
|
20 |
+
"B": 18,
|
21 |
+
"G": 19,
|
22 |
+
"V": 20,
|
23 |
+
"Y": 21,
|
24 |
+
"Q": 22,
|
25 |
+
"H": 23,
|
26 |
+
"F": 24,
|
27 |
+
"Z": 25,
|
28 |
+
"J": 26,
|
29 |
+
"Ñ": 27,
|
30 |
+
"X": 28,
|
31 |
+
"K": 29,
|
32 |
+
"W": 30
|
33 |
+
}
|