Training in progress, epoch 1

config.json

@@ -1,13 +1,11 @@
 {
-  "_name_or_path": "klue/roberta-base",
+  "_name_or_path": "kykim/electra-kor-base",
   "architectures": [
-    "BertForSequenceClassification"
+    "ElectraForSequenceClassification"
   ],
   "attention_probs_dropout_prob": 0.1,
-  "bos_token_id": 0,
   "classifier_dropout": null,
-  "eos_token_id": 2,
-  "gradient_checkpointing": false,
+  "embedding_size": 768,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
   "hidden_size": 768,
@@ -49,18 +47,21 @@
     "LABEL_8": 8,
     "LABEL_9": 9
   },
-  "layer_norm_eps": 1e-05,
-  "max_position_embeddings": 514,
-  "model_type": "bert",
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "electra",
   "num_attention_heads": 12,
   "num_hidden_layers": 12,
-  "pad_token_id": 1,
+  "pad_token_id": 0,
   "position_embedding_type": "absolute",
   "problem_type": "single_label_classification",
-  "tokenizer_class": "BertTokenizer",
+  "summary_activation": "gelu",
+  "summary_last_dropout": 0.1,
+  "summary_type": "first",
+  "summary_use_proj": true,
   "torch_dtype": "float32",
   "transformers_version": "4.33.2",
-  "type_vocab_size": 1,
+  "type_vocab_size": 2,
   "use_cache": true,
-  "vocab_size": 32000
+  "vocab_size": 42000
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:86e333d9f9d44bcf413478eb83bbc10fd959a4f2578937885951129b200e5ab6
-size 442590065
+oid sha256:eb5c70c73b9125c50e11dd9584e60ef233c40ffef02ea05946e1fb910f2549ca
+size 473307569

special_tokens_map.json

@@ -1,7 +1,5 @@
 {
-  "bos_token": "[CLS]",
   "cls_token": "[CLS]",
-  "eos_token": "[SEP]",
   "mask_token": "[MASK]",
   "pad_token": "[PAD]",
   "sep_token": "[SEP]",

tokenizer_config.json

@@ -1,16 +1,14 @@
 {
-  "bos_token": "[CLS]",
   "clean_up_tokenization_spaces": true,
   "cls_token": "[CLS]",
   "do_basic_tokenize": true,
-  "do_lower_case": false,
-  "eos_token": "[SEP]",
+  "do_lower_case": true,
   "mask_token": "[MASK]",
-  "model_max_length": 512,
+  "model_max_length": 256,
   "never_split": null,
   "pad_token": "[PAD]",
   "sep_token": "[SEP]",
-  "strip_accents": null,
+  "strip_accents": false,
   "tokenize_chinese_chars": true,
   "tokenizer_class": "BertTokenizer",
   "unk_token": "[UNK]"

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7f84d2049a84c6e2df076e3eaacaaf4bcb0c155033a15804ecb418da33aa8be3
+oid sha256:4b6bce61b6e553317a28affa96b556a06d071db455fa31cf2a63dc8f8ffef02c
 size 4091