Hugging Face's logo

santoshpandey
/
wav2vec2-large-xls-r-300m-nepali-colab

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
wav2vec2
Generated from Trainer
Model card Files Metrics Community
santoshpandey commited on
Commit
8ce7bcd
·
verified ·
1 Parent(s): 1d93adb

Upload tokenizer

Browse files
Files changed (3) hide show
  1. added_tokens.json +1 -1
  2. tokenizer_config.json +17 -1
  3. vocab.json +2 -2
added_tokens.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
  "</s>": 50,
3
- "[PAD]": 49
4
  }
 
1
  {
2
  "</s>": 50,
3
+ "<s>": 49
4
  }
tokenizer_config.json CHANGED
@@ -1,6 +1,14 @@
1
  {
2
  "added_tokens_decoder": {
3
- "49": {
 
 
 
 
 
 
 
 
4
  "content": "[PAD]",
5
  "lstrip": true,
6
  "normalized": false,
@@ -8,6 +16,14 @@
8
  "single_word": false,
9
  "special": false
10
  },
 
 
 
 
 
 
 
 
11
  "50": {
12
  "content": "</s>",
13
  "lstrip": false,
 
1
  {
2
  "added_tokens_decoder": {
3
+ "47": {
4
+ "content": "[UNK]",
5
+ "lstrip": true,
6
+ "normalized": false,
7
+ "rstrip": true,
8
+ "single_word": false,
9
+ "special": false
10
+ },
11
+ "48": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
 
16
  "single_word": false,
17
  "special": false
18
  },
19
+ "49": {
20
+ "content": "<s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
  "50": {
28
  "content": "</s>",
29
  "lstrip": false,
vocab.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
- "[PAD]": 49,
3
- "[UNK]": 49,
4
  "|": 0,
5
  "ँ": 1,
6
  "ं": 2,
 
1
  {
2
+ "[PAD]": 48,
3
+ "[UNK]": 47,
4
  "|": 0,
5
  "ँ": 1,
6
  "ं": 2,