Training in progress, step 100

config.json

@@ -1,36 +1,35 @@
 {
-  "_name_or_path": "EleutherAI/polyglot-ko-1.3b",
+  "_name_or_path": "klue/roberta-large",
   "architectures": [
-    "GPTNeoXForSequenceClassification"
+    "RobertaForSequenceClassification"
   ],
-  "attention_dropout": 0.0,
+  "attention_probs_dropout_prob": 0.1,
   "bos_token_id": 0,
-  "classifier_dropout": 0.1,
+  "classifier_dropout": null,
   "eos_token_id": 2,
+  "gradient_checkpointing": false,
   "hidden_act": "gelu",
-  "hidden_dropout": 0.0,
-  "hidden_size": 2048,
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
   "id2label": {
     "0": "LABEL_0"
   },
   "initializer_range": 0.02,
-  "intermediate_size": 8192,
+  "intermediate_size": 4096,
   "label2id": {
     "LABEL_0": 0
   },
   "layer_norm_eps": 1e-05,
-  "max_position_embeddings": 2048,
-  "model_type": "gpt_neox",
+  "max_position_embeddings": 514,
+  "model_type": "roberta",
   "num_attention_heads": 16,
   "num_hidden_layers": 24,
   "pad_token_id": 2,
-  "rope_scaling": null,
-  "rotary_emb_base": 10000,
-  "rotary_pct": 0.5,
-  "tie_word_embeddings": false,
+  "position_embedding_type": "absolute",
+  "tokenizer_class": "BertTokenizer",
   "torch_dtype": "float32",
   "transformers_version": "4.34.1",
+  "type_vocab_size": 1,
   "use_cache": true,
-  "use_parallel_residual": true,
-  "vocab_size": 30080
+  "vocab_size": 32000
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:04e0fdda0bca7be5e087fea60a849774815a7514862dc03d0f48c606db868886
-size 5080945162
+oid sha256:59587ddb0c8585b1b8154e3d30ecf54a9e18dae1ce009bcdaee1aac965f85089
+size 1346765678

special_tokens_map.json

@@ -1,11 +1,9 @@
 {
-  "additional_special_tokens": [
-    "<|endoftext|>",
-    "<|sep|>",
-    "<|acc|>",
-    "<|tel|>",
-    "<|rrn|>"
-  ],
-  "eos_token": "<|endoftext|>",
-  "pad_token": "<|endoftext|>"
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
 }

tokenizer_config.json

@@ -1,7 +1,7 @@
 {
   "added_tokens_decoder": {
     "0": {
-      "content": "<|unused0|>",
+      "content": "[CLS]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -9,7 +9,7 @@
       "special": true
     },
     "1": {
-      "content": "<|unused1|>",
+      "content": "[PAD]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -17,7 +17,7 @@
       "special": true
     },
     "2": {
-      "content": "<|endoftext|>",
+      "content": "[SEP]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -25,31 +25,15 @@
       "special": true
     },
     "3": {
-      "content": "<|sep|>",
+      "content": "[UNK]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
       "single_word": false,
       "special": true
     },
-    "30000": {
-      "content": "<|acc|>",
-      "lstrip": false,
-      "normalized": false,
-      "rstrip": false,
-      "single_word": false,
-      "special": true
-    },
-    "30001": {
-      "content": "<|tel|>",
-      "lstrip": false,
-      "normalized": false,
-      "rstrip": false,
-      "single_word": false,
-      "special": true
-    },
-    "30002": {
-      "content": "<|rrn|>",
+    "4": {
+      "content": "[MASK]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -57,16 +41,19 @@
       "special": true
     }
   },
-  "additional_special_tokens": [
-    "<|endoftext|>",
-    "<|sep|>",
-    "<|acc|>",
-    "<|tel|>",
-    "<|rrn|>"
-  ],
+  "bos_token": "[CLS]",
   "clean_up_tokenization_spaces": true,
-  "eos_token": "<|endoftext|>",
-  "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "<|endoftext|>",
-  "tokenizer_class": "PreTrainedTokenizerFast"
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "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 sha256:4f83b04586007121ba3ee6bfcd4a06860c6098884f4f89ada67554e532440c28
+oid sha256:e1b0be5f447301b577e44061c504ec9398fb05913a538169784e7cfc7134c7d3
 size 4600