Upload RobertaForMaskedLM

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "./models/06-12-17:35:49/checkpoint-47000",
+  "_name_or_path": "/opt/ml/level2_nlp_mrc-nlp-03/code/pretraining/models/06-17-06:11:12/checkpoint-55000",
   "architectures": [
     "RobertaForMaskedLM"
   ],
@@ -10,14 +10,14 @@
   "gradient_checkpointing": false,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
+  "hidden_size": 1024,
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
+  "intermediate_size": 4096,
   "layer_norm_eps": 1e-05,
   "max_position_embeddings": 514,
   "model_type": "roberta",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
   "pad_token_id": 1,
   "position_embedding_type": "absolute",
   "tokenizer_class": "BertTokenizer",

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb976ecdb0049ae10bd40f878d54024839ae0216899151a32b7fb0b336afbcc3
-size 442677497
+oid sha256:a7130ee4d2cd56a992a9b7cdf6b0781550a66f935d5f52eda2dfe0c40215047b
+size 1346896569