Upload tokenizer
Browse files- .gitattributes +2 -0
- concept/special_tokens_map.json +3 -0
- concept/tokenizer.json +0 -0
- concept/tokenizer_config.json +0 -0
- entity/special_tokens_map.json +3 -0
- entity/tokenizer.json +3 -0
- entity/tokenizer_config.json +0 -0
- item/special_tokens_map.json +4 -0
- item/tokenizer.json +3 -0
- item/tokenizer_config.json +0 -0
- tokenizer_config.json +9 -0
- word/special_tokens_map.json +3 -0
- word/tokenizer.json +0 -0
- word/tokenizer_config.json +0 -0
.gitattributes
CHANGED
@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
+
entity/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
37 |
+
item/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
concept/special_tokens_map.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"pad_token": "[UNK]"
|
3 |
+
}
|
concept/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
concept/tokenizer_config.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
entity/special_tokens_map.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"pad_token": "[PAD]"
|
3 |
+
}
|
entity/tokenizer.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0798d6ea50e1c1d335775ea7a54a659a841babd39203452b52c13ca62e4cef45
|
3 |
+
size 16653821
|
entity/tokenizer_config.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
item/special_tokens_map.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"pad_token": "[PAD]",
|
3 |
+
"unk_token": "[UNK]"
|
4 |
+
}
|
item/tokenizer.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5c5367496303729b85ae8f9329823d91a96a27ab30732735804a0b175494c43d
|
3 |
+
size 14121752
|
item/tokenizer_config.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"tokenizers": {
|
3 |
+
"word": "KGSFBaseTokenizer",
|
4 |
+
"item": "EntityTokenizer",
|
5 |
+
"entity": "KGSFBaseTokenizer",
|
6 |
+
"concept": "KGSFBaseTokenizer"
|
7 |
+
},
|
8 |
+
"tokenizer_key_for_decode": "word"
|
9 |
+
}
|
word/special_tokens_map.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"pad_token": "[PAD]"
|
3 |
+
}
|
word/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
word/tokenizer_config.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|