DAPT Kcbert

config.json

@@ -1,25 +1,31 @@
 {
-  "_name_or_path": "klue/bert-base",
+  "_name_or_path": "beomi/kcbert-base",
   "architectures": [
-    "BertForMaskedLM"
+    "BertForPreTraining"
   ],
   "attention_probs_dropout_prob": 0.1,
   "classifier_dropout": null,
+  "directionality": "bidi",
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
   "hidden_size": 768,
   "initializer_range": 0.02,
   "intermediate_size": 3072,
   "layer_norm_eps": 1e-12,
-  "max_position_embeddings": 512,
+  "max_position_embeddings": 300,
   "model_type": "bert",
   "num_attention_heads": 12,
   "num_hidden_layers": 12,
   "pad_token_id": 0,
+  "pooler_fc_size": 768,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
   "position_embedding_type": "absolute",
   "torch_dtype": "float32",
   "transformers_version": "4.24.0",
   "type_vocab_size": 2,
   "use_cache": true,
-  "vocab_size": 32000
+  "vocab_size": 30000
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:536d3edf06062b27b740d6b5b5322c2d493c974f6836427aa78bbf9dd1a3e14a
-size 442676985
+oid sha256:371c929aec905e84da26aab95c714411c2b31612d432e9bbe3a638a25d09a354
+size 438241801

tokenizer_config.json

@@ -3,12 +3,12 @@
   "do_basic_tokenize": true,
   "do_lower_case": false,
   "mask_token": "[MASK]",
-  "model_max_length": 512,
-  "name_or_path": "klue/bert-base",
+  "model_max_length": 300,
+  "name_or_path": "beomi/kcbert-base",
   "never_split": null,
   "pad_token": "[PAD]",
   "sep_token": "[SEP]",
-  "special_tokens_map_file": "/data1/hanyang/.cache/huggingface/hub/models--klue--bert-base/snapshots/34b965303f98bc5214daca7f76b7fb82d2dc6183/special_tokens_map.json",
+  "special_tokens_map_file": null,
   "strip_accents": null,
   "tokenize_chinese_chars": true,
   "tokenizer_class": "BertTokenizer",