Revert: clean tokenizer_config matching training IDs

tokenizer_config.json

@@ -1,4 +1,17 @@
 {
+  "add_prefix_space": true,
+  "backend": "custom",
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": [],
+  "mask_token": "<mask>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>",
+  "unk_id": 1,
+  "tokenizer_class": "XLMRobertaTokenizer",
   "added_tokens_decoder": {
     "0": {
       "content": "<pad>",
@@ -34,20 +47,11 @@
     },
     "32000": {
       "content": "<mask>",
-      "lstrip": false,
+      "lstrip": true,
       "normalized": false,
       "rstrip": false,
       "single_word": false,
       "special": true
     }
-  },
-  "additional_special_tokens": null,
-  "backend": "custom",
-  "bos_token": "<s>",
-  "eos_token": "</s>",
-  "mask_token": "<mask>",
-  "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "<pad>",
-  "tokenizer_class": "XLMRobertaTokenizer",
-  "unk_token": "<unk>"
+  }
 }