Training in progress, epoch 1

config.json

@@ -1,14 +1,13 @@
 {
-  "_name_or_path": "bert-base-cased",
+  "_name_or_path": "distilbert-base-cased",
+  "activation": "gelu",
   "architectures": [
-    "BertForTokenClassification"
+    "DistilBertForTokenClassification"
   ],
-  "attention_probs_dropout_prob": 0.1,
-  "classifier_dropout": null,
-  "gradient_checkpointing": false,
-  "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
+  "attention_dropout": 0.1,
+  "dim": 768,
+  "dropout": 0.1,
+  "hidden_dim": 3072,
   "id2label": {
     "0": "O",
     "1": "B-Age",
@@ -58,7 +57,6 @@
     "45": "B-Outcome"
   },
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
   "label2id": {
     "B-Administration": 21,
     "B-Age": 1,
@@ -107,16 +105,17 @@
     "I-Time": 36,
     "O": 0
   },
-  "layer_norm_eps": 1e-12,
   "max_position_embeddings": 512,
-  "model_type": "bert",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
+  "model_type": "distilbert",
+  "n_heads": 12,
+  "n_layers": 6,
+  "output_past": true,
   "pad_token_id": 0,
-  "position_embedding_type": "absolute",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
   "torch_dtype": "float32",
   "transformers_version": "4.29.2",
-  "type_vocab_size": 2,
-  "use_cache": true,
   "vocab_size": 28996
 }

label2id.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:294b06dff14c32d1ad668f2c8ea9cffced02c9f5ebfff05b38bbf53560a0203b
-size 1549
+oid sha256:8bd811dceffbe8543de7c82ed5fa17028251aa639bfec0a867cabe0868e79070
+size 853

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ccfe89c3401223569a09c578c20a919f7c9b87ab33828900e5b7dfe5af900e56
-size 431092333
+oid sha256:12d94bba34705f450cadfbaac174405cb19f822fda48f1bccec6fec239d4a694
+size 260939941

runs/Jun18_19-11-32_u116734/1687140697.0187914/events.out.tfevents.1687140697.u116734.1156936.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b097a9b7400e677d92738bec666a5b321a226efcf9a0840d004964f6e544470
+size 5902

runs/Jun18_19-11-32_u116734/events.out.tfevents.1687140697.u116734.1156936.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:239d378692fdd99feb1df3809683514fe77662a10af4476508709ee2af75287f
+size 6503

tokenizer.json

@@ -6,7 +6,14 @@
     "strategy": "LongestFirst",
     "stride": 0
   },
-  "padding": null,
+  "padding": {
+    "strategy": "BatchLongest",
+    "direction": "Right",
+    "pad_to_multiple_of": null,
+    "pad_id": 0,
+    "pad_type_id": 0,
+    "pad_token": "[PAD]"
+  },
   "added_tokens": [
     {
       "id": 0,

tokenizer_config.json

@@ -8,6 +8,6 @@
   "sep_token": "[SEP]",
   "strip_accents": null,
   "tokenize_chinese_chars": true,
-  "tokenizer_class": "BertTokenizer",
+  "tokenizer_class": "DistilBertTokenizer",
   "unk_token": "[UNK]"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b8d9fe341e88470ae92fa16e68eb47fab0ad970d29e92601908ed5e3a56d42b7
+oid sha256:69eb03bbd2a99fb3fdce50e03a573b236ac6785fc293dd38cfb0e08db09c2962
 size 3899