fix: update tokenizer_config

tokenizer_config.json

@@ -1,5 +1,5 @@
 {
-  "add_prefix_space": false,
+  "add_prefix_space": true,
   "added_tokens_decoder": {
     "50256": {
       "content": "<|endoftext|>",
@@ -15,5 +15,6 @@
   "eos_token": "<|endoftext|>",
   "model_max_length": 1024,
   "tokenizer_class": "GPT2Tokenizer",
-  "unk_token": "<|endoftext|>"
-}
+  "unk_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>",
+}