Upload tokenizer

special_tokens_map.json

@@ -13,7 +13,13 @@
     "rstrip": false,
     "single_word": false
   },
-  "pad_token": "#",
+  "pad_token": {
+    "content": "\"",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
   "unk_token": {
     "content": "<|endoftext|>",
     "lstrip": false,

tokenizer.json

@@ -58,6 +58,7 @@
     "end_of_word_suffix": "",
     "fuse_unk": false,
     "byte_fallback": false,
+    "ignore_merges": false,
     "vocab": {
       "!": 0,
       "\"": 1,

tokenizer_config.json

@@ -31,8 +31,9 @@
   "clean_up_tokenization_spaces": true,
   "eos_token": "#",
   "errors": "replace",
-  "model_max_length": 1024,
-  "pad_token": "#",
+  "model_max_length": 2048,
+  "pad_token": "\"",
+  "padding_side": "left",
   "tokenizer_class": "GPT2Tokenizer",
   "unk_token": "<|endoftext|>"
 }