add tokenizer

special_tokens_map.json

@@ -1,4 +1,16 @@
 {
+  "additional_special_tokens": [
+    "number0",
+    "number1",
+    "number2",
+    "number3",
+    "number4",
+    "number5",
+    "number6",
+    "number7",
+    "number8",
+    "number9"
+  ],
   "bos_token": "<s>",
   "eos_token": "</s>",
   "pad_token": "<pad>",

tokenizer.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6c1abd73425d69d27b6933af4fa2a004568434169689b37d1314c6ca3a1d2a7f
-size 14500541
+oid sha256:98b43f068807538425855716f1fb28c325234defb3f651ab3513dd37ba4ffb03
+size 14502381

tokenizer_config.json

@@ -1,5 +1,18 @@
 {
   "add_prefix_space": false,
+  "add_special_tokens": true,
+  "additional_special_tokens": [
+    "number0",
+    "number1",
+    "number2",
+    "number3",
+    "number4",
+    "number5",
+    "number6",
+    "number7",
+    "number8",
+    "number9"
+  ],
   "bos_token": "<s>",
   "eos_token": "</s>",
   "name_or_path": "bigscience/bloom-560m",