Training in progress, step 3500

config.json

@@ -1,17 +1,13 @@
 {
-  "_name_or_path": "albert-base-v2",
+  "_name_or_path": "bert-base-uncased",
   "architectures": [
-    "AlbertForSequenceClassification"
+    "BertForSequenceClassification"
   ],
-  "attention_probs_dropout_prob": 0,
-  "bos_token_id": 2,
-  "classifier_dropout_prob": 0.1,
-  "down_scale_factor": 1,
-  "embedding_size": 128,
-  "eos_token_id": 3,
-  "gap_size": 0,
-  "hidden_act": "gelu_new",
-  "hidden_dropout_prob": 0,
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
   "hidden_size": 768,
   "id2label": {
     "0": "LABEL_0",
@@ -22,7 +18,6 @@
     "5": "LABEL_5"
   },
   "initializer_range": 0.02,
-  "inner_group_num": 1,
   "intermediate_size": 3072,
   "label2id": {
     "LABEL_0": 0,
@@ -34,17 +29,15 @@
   },
   "layer_norm_eps": 1e-12,
   "max_position_embeddings": 512,
-  "model_type": "albert",
-  "net_structure_type": 0,
+  "model_type": "bert",
   "num_attention_heads": 12,
-  "num_hidden_groups": 1,
   "num_hidden_layers": 12,
-  "num_memory_blocks": 0,
   "pad_token_id": 0,
   "position_embedding_type": "absolute",
   "problem_type": "single_label_classification",
   "torch_dtype": "float32",
   "transformers_version": "4.38.2",
   "type_vocab_size": 2,
-  "vocab_size": 30000
+  "use_cache": true,
+  "vocab_size": 30522
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:63343fb19b8c5a429e0565bc5b090ab31b0cdb15f95b524acf7edfecd5c10752
-size 46756216
+oid sha256:9c07d56cd7371f60e807f46c0fd6b50277b0c17bc6f9765c6be48e633effff29
+size 437970952

runs/Mar11_02-44-10_53107997d4b5/events.out.tfevents.1710125050.53107997d4b5.7340.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e3ee1037a225925bb670e8b58481ef2071a73026576bd30cf288933bbd8b2e19
-size 6862
+oid sha256:3b66b8725956719f7a36ea95d226e10e22803000e45c872c0903eeb44d7c6f16
+size 7396

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f7341803b24e9f8352aa2346b85e246be75ca1f3bc0be5ae2dd30dab8a232c7e
+oid sha256:b0e82a385dc85b5a76f530c7e140bfeced48c2fca249d832ecce6bee5c8d712c
 size 4920