Samuael
/

geez-asr

Automatic Speech Recognition

Generated from Trainer

Inference Endpoints

Model card Files Files and versions Metrics Training metrics Community

Samuael commited on Sep 21

Commit

e9caaab

•

1 Parent(s): e3b2b9b

Upload tokenizer

Files changed (3) hide show

added_tokens.json +2 -4
tokenizer_config.json +12 -12
vocab.json +36 -34

added_tokens.json CHANGED Viewed

@@ -1,6 +1,4 @@
 {
-  "</s>": 38,
-  "<s>": 37,
-  "[PAD]": 40,
-  "[UNK]": 39
 }

 {
+  "</s>": 40,
+  "<s>": 39
 }

tokenizer_config.json CHANGED Viewed

@@ -1,23 +1,23 @@
 {
   "added_tokens_decoder": {
-    "37": {
-      "content": "<s>",
-      "lstrip": false,
       "normalized": false,
-      "rstrip": false,
       "single_word": false,
-      "special": true
     },
-    "38": {
-      "content": "</s>",
-      "lstrip": false,
       "normalized": false,
-      "rstrip": false,
       "single_word": false,
-      "special": true
     },
     "39": {
-      "content": "[UNK]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -25,7 +25,7 @@
       "special": true
     },
     "40": {
-      "content": "[PAD]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,

 {
   "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": true,
       "normalized": false,
+      "rstrip": true,
       "single_word": false,
+      "special": false
     },
+    "1": {
+      "content": "[UNK]",
+      "lstrip": true,
       "normalized": false,
+      "rstrip": true,
       "single_word": false,
+      "special": false
     },
     "39": {
+      "content": "<s>",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
       "special": true
     },
     "40": {
+      "content": "</s>",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,

vocab.json CHANGED Viewed

@@ -1,39 +1,41 @@
 {
-  " ": 16,
-  "ል": 19,
-  "ሕ": 9,
-  "ም": 33,
-  "ር": 24,
-  "ስ": 23,
-  "ሽ": 26,
-  "ቅ": 21,
-  "ቕ": 14,
-  "ብ": 25,
-  "ቭ": 6,
-  "ት": 17,
-  "ች": 2,
-  "ን": 20,
-  "ኝ": 34,
-  "ኡ": 31,
-  "ኢ": 12,
-  "ኣ": 29,
-  "ኤ": 5,
   "እ": 22,
-  "ኦ": 27,
-  "ኧ": 18,
-  "ክ": 30,
-  "ኽ": 28,
-  "ው": 10,
   "ዝ": 7,
-  "ዥ": 1,
-  "ይ": 8,
-  "ድ": 15,
-  "ጅ": 35,
-  "ግ": 36,
-  "ጥ": 0,
-  "ጭ": 13,
   "ጵ": 11,
-  "ጽ": 3,
-  "ፍ": 32,
-  "ፕ": 4
 }

 {
+  "[PAD]": 0,
+  "[UNK]": 1,
+  "|": 2,
+  "ል": 37,
+  "ሕ": 30,
+  "ም": 13,
+  "ር": 8,
+  "ስ": 17,
+  "ሽ": 3,
+  "ቅ": 26,
+  "ቕ": 9,
+  "ብ": 29,
+  "ቭ": 35,
+  "ት": 34,
+  "ች": 20,
+  "ን": 36,
+  "ኝ": 27,
+  "ኡ": 21,
+  "ኢ": 31,
+  "ኣ": 14,
+  "ኤ": 32,
   "እ": 22,
+  "ኦ": 12,
+  "ኧ": 15,
+  "ክ": 28,
+  "ኽ": 6,
+  "ው": 25,
   "ዝ": 7,
+  "ዥ": 4,
+  "ይ": 23,
+  "ድ": 5,
+  "ጅ": 18,
+  "ግ": 38,
+  "ጥ": 10,
+  "ጭ": 33,
   "ጵ": 11,
+  "ጽ": 16,
+  "ፍ": 19,
+  "ፕ": 24
 }