Commit
·
c3a8fbf
1
Parent(s):
627d437
Upload tokenizer
Browse files- added_tokens.json +5 -0
- special_tokens_map.json +3 -3
- tokenizer.model +2 -2
- tokenizer_config.json +4 -4
added_tokens.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<|bos|>": 52928,
|
| 3 |
+
"<|eos|>": 52929,
|
| 4 |
+
"<|pad|>": 52930
|
| 5 |
+
}
|
special_tokens_map.json
CHANGED
|
@@ -1,20 +1,20 @@
|
|
| 1 |
{
|
| 2 |
"bos_token": {
|
| 3 |
-
"content": "
|
| 4 |
"lstrip": false,
|
| 5 |
"normalized": true,
|
| 6 |
"rstrip": false,
|
| 7 |
"single_word": false
|
| 8 |
},
|
| 9 |
"eos_token": {
|
| 10 |
-
"content": "
|
| 11 |
"lstrip": false,
|
| 12 |
"normalized": true,
|
| 13 |
"rstrip": false,
|
| 14 |
"single_word": false
|
| 15 |
},
|
| 16 |
"pad_token": {
|
| 17 |
-
"content": "
|
| 18 |
"lstrip": false,
|
| 19 |
"normalized": true,
|
| 20 |
"rstrip": false,
|
|
|
|
| 1 |
{
|
| 2 |
"bos_token": {
|
| 3 |
+
"content": "<|bos|>",
|
| 4 |
"lstrip": false,
|
| 5 |
"normalized": true,
|
| 6 |
"rstrip": false,
|
| 7 |
"single_word": false
|
| 8 |
},
|
| 9 |
"eos_token": {
|
| 10 |
+
"content": "<|eos|>",
|
| 11 |
"lstrip": false,
|
| 12 |
"normalized": true,
|
| 13 |
"rstrip": false,
|
| 14 |
"single_word": false
|
| 15 |
},
|
| 16 |
"pad_token": {
|
| 17 |
+
"content": "<|pad|>",
|
| 18 |
"lstrip": false,
|
| 19 |
"normalized": true,
|
| 20 |
"rstrip": false,
|
tokenizer.model
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:68d82608100430846be804d586a586a80fc77d4f0eb06ee6d62133a47c9beb9f
|
| 3 |
+
size 827607
|
tokenizer_config.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"add_bos_token": true,
|
| 3 |
-
"add_eos_token":
|
| 4 |
"bos_token": {
|
| 5 |
"__type": "AddedToken",
|
| 6 |
-
"content": "
|
| 7 |
"lstrip": false,
|
| 8 |
"normalized": true,
|
| 9 |
"rstrip": false,
|
|
@@ -12,7 +12,7 @@
|
|
| 12 |
"clean_up_tokenization_spaces": false,
|
| 13 |
"eos_token": {
|
| 14 |
"__type": "AddedToken",
|
| 15 |
-
"content": "
|
| 16 |
"lstrip": false,
|
| 17 |
"normalized": true,
|
| 18 |
"rstrip": false,
|
|
@@ -21,7 +21,7 @@
|
|
| 21 |
"model_max_length": 1000000000000000019884624838656,
|
| 22 |
"pad_token": {
|
| 23 |
"__type": "AddedToken",
|
| 24 |
-
"content": "
|
| 25 |
"lstrip": false,
|
| 26 |
"normalized": true,
|
| 27 |
"rstrip": false,
|
|
|
|
| 1 |
{
|
| 2 |
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": true,
|
| 4 |
"bos_token": {
|
| 5 |
"__type": "AddedToken",
|
| 6 |
+
"content": "<|bos|>",
|
| 7 |
"lstrip": false,
|
| 8 |
"normalized": true,
|
| 9 |
"rstrip": false,
|
|
|
|
| 12 |
"clean_up_tokenization_spaces": false,
|
| 13 |
"eos_token": {
|
| 14 |
"__type": "AddedToken",
|
| 15 |
+
"content": "<|eos|>",
|
| 16 |
"lstrip": false,
|
| 17 |
"normalized": true,
|
| 18 |
"rstrip": false,
|
|
|
|
| 21 |
"model_max_length": 1000000000000000019884624838656,
|
| 22 |
"pad_token": {
|
| 23 |
"__type": "AddedToken",
|
| 24 |
+
"content": "<|pad|>",
|
| 25 |
"lstrip": false,
|
| 26 |
"normalized": true,
|
| 27 |
"rstrip": false,
|