Upload tokenizer

README.md

@@ -1,10 +1,10 @@
 ---
-base_model: Samuael/amBART_1000
 tags:
 - generated_from_trainer
 metrics:
 - wer
 - bleu
+base_model: Samuael/amBART_1000
 model-index:
 - name: amBART_261
   results: []

sentencepiece.bpe.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f4e4c09bd68c20916dfa8472b3c979527234d7b22afa71e1ea0bb36ee79a1bbd
-size 253571
+oid sha256:a801c63cf0822cc3a880177fd5895196337d7e3813edde88c428061c263354a4
+size 240461

special_tokens_map.json

@@ -1,71 +1,12 @@
 {
   "additional_special_tokens": [
-    "ar_AR",
-    "cs_CZ",
-    "de_DE",
-    "en_XX",
-    "es_XX",
-    "et_EE",
-    "fi_FI",
-    "fr_XX",
-    "gu_IN",
-    "hi_IN",
-    "it_IT",
-    "ja_XX",
-    "kk_KZ",
-    "ko_KR",
-    "lt_LT",
-    "lv_LV",
-    "my_MM",
-    "ne_NP",
-    "nl_XX",
-    "ro_RO",
-    "ru_RU",
-    "si_LK",
-    "tr_TR",
-    "vi_VN",
-    "zh_CN"
+    "ቊ",
+    "ar_AR"
   ],
-  "bos_token": {
-    "content": "<s>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "cls_token": {
-    "content": "<s>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "eos_token": {
-    "content": "</s>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "pad_token": {
-    "content": "<pad>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "sep_token": {
-    "content": "</s>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "unk_token": {
-    "content": "<unk>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  }
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
 }

tokenizer_config.json

@@ -32,7 +32,7 @@
       "single_word": false,
       "special": true
     },
-    "1001": {
+    "235": {
       "content": "ar_AR",
       "lstrip": false,
       "normalized": false,
@@ -40,7 +40,7 @@
       "single_word": false,
       "special": true
     },
-    "1002": {
+    "236": {
       "content": "cs_CZ",
       "lstrip": false,
       "normalized": false,
@@ -48,7 +48,7 @@
       "single_word": false,
       "special": true
     },
-    "1003": {
+    "237": {
       "content": "de_DE",
       "lstrip": false,
       "normalized": false,
@@ -56,7 +56,7 @@
       "single_word": false,
       "special": true
     },
-    "1004": {
+    "238": {
       "content": "en_XX",
       "lstrip": false,
       "normalized": false,
@@ -64,7 +64,7 @@
       "single_word": false,
       "special": true
     },
-    "1005": {
+    "239": {
       "content": "es_XX",
       "lstrip": false,
       "normalized": false,
@@ -72,7 +72,7 @@
       "single_word": false,
       "special": true
     },
-    "1006": {
+    "240": {
       "content": "et_EE",
       "lstrip": false,
       "normalized": false,
@@ -80,7 +80,7 @@
       "single_word": false,
       "special": true
     },
-    "1007": {
+    "241": {
       "content": "fi_FI",
       "lstrip": false,
       "normalized": false,
@@ -88,7 +88,7 @@
       "single_word": false,
       "special": true
     },
-    "1008": {
+    "242": {
       "content": "fr_XX",
       "lstrip": false,
       "normalized": false,
@@ -96,7 +96,7 @@
       "single_word": false,
       "special": true
     },
-    "1009": {
+    "243": {
       "content": "gu_IN",
       "lstrip": false,
       "normalized": false,
@@ -104,7 +104,7 @@
       "single_word": false,
       "special": true
     },
-    "1010": {
+    "244": {
       "content": "hi_IN",
       "lstrip": false,
       "normalized": false,
@@ -112,7 +112,7 @@
       "single_word": false,
       "special": true
     },
-    "1011": {
+    "245": {
       "content": "it_IT",
       "lstrip": false,
       "normalized": false,
@@ -120,7 +120,7 @@
       "single_word": false,
       "special": true
     },
-    "1012": {
+    "246": {
       "content": "ja_XX",
       "lstrip": false,
       "normalized": false,
@@ -128,7 +128,7 @@
       "single_word": false,
       "special": true
     },
-    "1013": {
+    "247": {
       "content": "kk_KZ",
       "lstrip": false,
       "normalized": false,
@@ -136,7 +136,7 @@
       "single_word": false,
       "special": true
     },
-    "1014": {
+    "248": {
       "content": "ko_KR",
       "lstrip": false,
       "normalized": false,
@@ -144,7 +144,7 @@
       "single_word": false,
       "special": true
     },
-    "1015": {
+    "249": {
       "content": "lt_LT",
       "lstrip": false,
       "normalized": false,
@@ -152,7 +152,7 @@
       "single_word": false,
       "special": true
     },
-    "1016": {
+    "250": {
       "content": "lv_LV",
       "lstrip": false,
       "normalized": false,
@@ -160,7 +160,7 @@
       "single_word": false,
       "special": true
     },
-    "1017": {
+    "251": {
       "content": "my_MM",
       "lstrip": false,
       "normalized": false,
@@ -168,7 +168,7 @@
       "single_word": false,
       "special": true
     },
-    "1018": {
+    "252": {
       "content": "ne_NP",
       "lstrip": false,
       "normalized": false,
@@ -176,7 +176,7 @@
       "single_word": false,
       "special": true
     },
-    "1019": {
+    "253": {
       "content": "nl_XX",
       "lstrip": false,
       "normalized": false,
@@ -184,7 +184,7 @@
       "single_word": false,
       "special": true
     },
-    "1020": {
+    "254": {
       "content": "ro_RO",
       "lstrip": false,
       "normalized": false,
@@ -192,7 +192,7 @@
       "single_word": false,
       "special": true
     },
-    "1021": {
+    "255": {
       "content": "ru_RU",
       "lstrip": false,
       "normalized": false,
@@ -200,7 +200,7 @@
       "single_word": false,
       "special": true
     },
-    "1022": {
+    "256": {
       "content": "si_LK",
       "lstrip": false,
       "normalized": false,
@@ -208,7 +208,7 @@
       "single_word": false,
       "special": true
     },
-    "1023": {
+    "257": {
       "content": "tr_TR",
       "lstrip": false,
       "normalized": false,
@@ -216,7 +216,7 @@
       "single_word": false,
       "special": true
     },
-    "1024": {
+    "258": {
       "content": "vi_VN",
       "lstrip": false,
       "normalized": false,
@@ -224,7 +224,7 @@
       "single_word": false,
       "special": true
     },
-    "1025": {
+    "259": {
       "content": "zh_CN",
       "lstrip": false,
       "normalized": false,
@@ -234,31 +234,8 @@
     }
   },
   "additional_special_tokens": [
-    "ar_AR",
-    "cs_CZ",
-    "de_DE",
-    "en_XX",
-    "es_XX",
-    "et_EE",
-    "fi_FI",
-    "fr_XX",
-    "gu_IN",
-    "hi_IN",
-    "it_IT",
-    "ja_XX",
-    "kk_KZ",
-    "ko_KR",
-    "lt_LT",
-    "lv_LV",
-    "my_MM",
-    "ne_NP",
-    "nl_XX",
-    "ro_RO",
-    "ru_RU",
-    "si_LK",
-    "tr_TR",
-    "vi_VN",
-    "zh_CN"
+    "ቊ",
+    "ar_AR"
   ],
   "bos_token": "<s>",
   "clean_up_tokenization_spaces": true,
@@ -272,5 +249,6 @@
   "src_lang": "ar_AR",
   "tgt_lang": "cs_CZ",
   "tokenizer_class": "MBartTokenizer",
+  "tokenizer_file": null,
   "unk_token": "<unk>"
 }