Upload tokenizer

added_tokens.json

@@ -98,6 +98,5 @@
   "<extra_id_97>": 1002,
   "<extra_id_98>": 1001,
   "<extra_id_99>": 1000,
-  "<extra_id_9>": 1090,
-  "<pad>": 1100
+  "<extra_id_9>": 1090
 }

special_tokens_map.json

@@ -101,6 +101,7 @@
     "<extra_id_98>",
     "<extra_id_99>"
   ],
+  "bos_token": "<s>",
   "eos_token": {
     "content": "</s>",
     "lstrip": false,

spiece.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f4e4c09bd68c20916dfa8472b3c979527234d7b22afa71e1ea0bb36ee79a1bbd
-size 253571
+oid sha256:d862dc059c77d41f3108bbbd7578cc6d4f80fe3306247647c77c6747a39ef717
+size 253584

tokenizer_config.json

@@ -2,6 +2,14 @@
   "add_prefix_space": true,
   "added_tokens_decoder": {
     "0": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
       "content": "<unk>",
       "lstrip": false,
       "normalized": false,
@@ -10,6 +18,14 @@
       "special": true
     },
     "2": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
       "content": "</s>",
       "lstrip": false,
       "normalized": false,
@@ -816,14 +832,6 @@
       "rstrip": true,
       "single_word": false,
       "special": true
-    },
-    "1100": {
-      "content": "<pad>",
-      "lstrip": false,
-      "normalized": false,
-      "rstrip": false,
-      "single_word": false,
-      "special": true
     }
   },
   "additional_special_tokens": [
@@ -928,12 +936,16 @@
     "<extra_id_98>",
     "<extra_id_99>"
   ],
+  "bos_token": "<s>",
   "clean_up_tokenization_spaces": true,
+  "do_basic_tokenize": true,
+  "do_lower_case": true,
   "eos_token": "</s>",
   "extra_ids": 100,
   "legacy": true,
   "model_max_length": 1000000000000000019884624838656,
   "pad_token": "<pad>",
+  "padding": true,
   "sp_model_kwargs": {},
   "tokenizer_class": "T5Tokenizer",
   "unk_token": "<unk>"