update tokenizer

README.md

@@ -1,5 +1,5 @@
 ---
-pipeline_tag: feature-extraction
+pipeline_tag: sentence-similarity
 language: ja
 license: cc-by-sa-4.0
 tags:

README_JA.md

@@ -1,5 +1,5 @@
 ---
-pipeline_tag: feature-extraction
+pipeline_tag: sentence-similarity
 language: ja
 license: cc-by-sa-4.0
 tags:

config.json

@@ -1,10 +1,8 @@
 {
-  "_name_or_path": "cl-tohoku/bert-base-japanese-v2",
   "architectures": [
     "BertModel"
   ],
   "attention_probs_dropout_prob": 0.1,
-  "classifier_dropout": null,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
   "hidden_size": 768,
@@ -16,11 +14,7 @@
   "num_attention_heads": 12,
   "num_hidden_layers": 12,
   "pad_token_id": 0,
-  "position_embedding_type": "absolute",
   "tokenizer_class": "BertJapaneseTokenizer",
-  "torch_dtype": "float32",
-  "transformers_version": "4.25.1",
   "type_vocab_size": 2,
-  "use_cache": true,
   "vocab_size": 32768
 }

requirements.txt

@@ -1 +0,0 @@
-fugashi[unidic-lite]

special_tokens_map.json

@@ -1,7 +0,0 @@
-{
-  "cls_token": "[CLS]",
-  "mask_token": "[MASK]",
-  "pad_token": "[PAD]",
-  "sep_token": "[SEP]",
-  "unk_token": "[UNK]"
-}

tokenizer_config.json

@@ -2,19 +2,17 @@
   "do_lower_case": false,
   "do_subword_tokenize": true,
   "do_word_tokenize": true,
-  "jumanpp_kwargs": null,
   "mask_token": "[MASK]",
   "mecab_kwargs": {
     "mecab_dic": "unidic_lite"
   },
   "model_max_length": 512,
-  "name_or_path": "cl-tohoku/bert-base-japanese-v2",
-  "never_split": null,
   "pad_token": "[PAD]",
+  "cls_token": "[CLS]",
   "sep_token": "[SEP]",
   "special_tokens_map_file": null,
+  "name_or_path": "cl-tohoku/bert-base-japanese-v2",
   "subword_tokenizer_type": "wordpiece",
-  "mecab_kwargs": {
-      "mecab_dic": "unidic_lite"
-  }
+  "unk_token": "[UNK]",
+  "word_tokenizer_type": "mecab"
 }