hadiqa123
/

xls-r_cv_ur

Automatic Speech Recognition

Generated from Trainer

Inference Endpoints

Model card Files Files and versions Metrics Training metrics Community

hadiqa123 commited on Jul 29, 2023

Commit

7584cef

•

1 Parent(s): 03066fc

Upload tokenizer

Files changed (4) hide show

added_tokens.json +2 -2
special_tokens_map.json +16 -0
tokenizer_config.json +3 -0
vocab.json +44 -43

added_tokens.json CHANGED Viewed

@@ -1,4 +1,4 @@
 {
-  "</s>": 46,
-  "<s>": 45
 }

 {
+  "</s>": 47,
+  "<s>": 46
 }

special_tokens_map.json CHANGED Viewed

@@ -1,4 +1,20 @@
 {
   "bos_token": "<s>",
   "eos_token": "</s>",
   "pad_token": "[PAD]",

 {
+  "additional_special_tokens": [
+    {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false
+    },
+    {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false
+    }
+  ],
   "bos_token": "<s>",
   "eos_token": "</s>",
   "pad_token": "[PAD]",

tokenizer_config.json CHANGED Viewed

@@ -1,9 +1,12 @@
 {
   "bos_token": "<s>",
   "do_lower_case": false,
   "eos_token": "</s>",
   "pad_token": "[PAD]",
   "replace_word_delimiter_char": " ",
   "tokenizer_class": "Wav2Vec2CTCTokenizer",
   "unk_token": "[UNK]",
   "word_delimiter_token": "|"

 {
   "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
   "do_lower_case": false,
   "eos_token": "</s>",
+  "model_max_length": 1000000000000000019884624838656,
   "pad_token": "[PAD]",
   "replace_word_delimiter_char": " ",
+  "target_lang": null,
   "tokenizer_class": "Wav2Vec2CTCTokenizer",
   "unk_token": "[UNK]",
   "word_delimiter_token": "|"

vocab.json CHANGED Viewed

@@ -1,47 +1,48 @@
 {
-  " ": 35,
-  "PAD": 43,
   "[UNK]": 44,
-  "آ": 7,
-  "أ": 2,
-  "ؤ": 27,
-  "ئ": 15,
-  "ا": 26,
-  "ب": 39,
-  "ت": 32,
-  "ث": 31,
-  "ج": 11,
-  "ح": 29,
-  "خ": 37,
-  "د": 23,
-  "ذ": 21,
-  "ر": 34,
-  "ز": 20,
-  "س": 18,
-  "ش": 33,
-  "ص": 42,
-  "ض": 5,
-  "ط": 9,
-  "ظ": 8,
-  "ع": 24,
-  "غ": 28,
-  "ف": 0,
   "ق": 22,
-  "ل": 19,
-  "م": 4,
-  "ن": 3,
-  "و": 14,
-  "ً": 13,
-  "ٹ": 38,
-  "پ": 36,
-  "چ": 6,
-  "ڈ": 17,
-  "ڑ": 40,
-  "ک": 16,
-  "گ": 12,
-  "ں": 30,
-  "ھ": 25,
-  "ہ": 10,
-  "ی": 1,
-  "ے": 41
 }

 {
+  "[PAD]": 45,
   "[UNK]": 44,
+  "|": 0,
+  "ء": 1,
+  "ا": 2,
+  "ب": 3,
+  "ت": 4,
+  "ث": 5,
+  "ج": 6,
+  "ح": 7,
+  "خ": 8,
+  "د": 9,
+  "ذ": 10,
+  "ر": 11,
+  "ز": 12,
+  "س": 13,
+  "ش": 14,
+  "ص": 15,
+  "ض": 16,
+  "ط": 17,
+  "ظ": 18,
+  "ع": 19,
+  "غ": 20,
+  "ف": 21,
   "ق": 22,
+  "ل": 23,
+  "م": 24,
+  "ن": 25,
+  "و": 26,
+  "ى": 27,
+  "ي": 28,
+  "ً": 29,
+  "ٓ": 30,
+  "ٹ": 31,
+  "پ": 32,
+  "چ": 33,
+  "ڈ": 34,
+  "ڑ": 35,
+  "ژ": 36,
+  "ک": 37,
+  "گ": 38,
+  "ں": 39,
+  "ھ": 40,
+  "ہ": 41,
+  "ی": 42,
+  "ے": 43
 }