Train 10k + CRF for financial-pii-ready. F1: 1.0000
Browse files
financial-pii-ready/TokenBased-BERT/config.json
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"directionality": "bidi",
|
| 8 |
+
"dtype": "float32",
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "B-ADDRESS",
|
| 14 |
+
"1": "B-COMPANY",
|
| 15 |
+
"2": "B-CREDIT_CARD",
|
| 16 |
+
"3": "B-EMAIL",
|
| 17 |
+
"4": "B-NAME",
|
| 18 |
+
"5": "B-PHONE",
|
| 19 |
+
"6": "B-SSN/CCCD",
|
| 20 |
+
"7": "B-URL",
|
| 21 |
+
"8": "I-ADDRESS",
|
| 22 |
+
"9": "I-COMPANY",
|
| 23 |
+
"10": "I-CREDIT_CARD",
|
| 24 |
+
"11": "I-EMAIL",
|
| 25 |
+
"12": "I-NAME",
|
| 26 |
+
"13": "I-PHONE",
|
| 27 |
+
"14": "I-SSN/CCCD",
|
| 28 |
+
"15": "I-URL",
|
| 29 |
+
"16": "O"
|
| 30 |
+
},
|
| 31 |
+
"initializer_range": 0.02,
|
| 32 |
+
"intermediate_size": 3072,
|
| 33 |
+
"label2id": {
|
| 34 |
+
"B-ADDRESS": 0,
|
| 35 |
+
"B-COMPANY": 1,
|
| 36 |
+
"B-CREDIT_CARD": 2,
|
| 37 |
+
"B-EMAIL": 3,
|
| 38 |
+
"B-NAME": 4,
|
| 39 |
+
"B-PHONE": 5,
|
| 40 |
+
"B-SSN/CCCD": 6,
|
| 41 |
+
"B-URL": 7,
|
| 42 |
+
"I-ADDRESS": 8,
|
| 43 |
+
"I-COMPANY": 9,
|
| 44 |
+
"I-CREDIT_CARD": 10,
|
| 45 |
+
"I-EMAIL": 11,
|
| 46 |
+
"I-NAME": 12,
|
| 47 |
+
"I-PHONE": 13,
|
| 48 |
+
"I-SSN/CCCD": 14,
|
| 49 |
+
"I-URL": 15,
|
| 50 |
+
"O": 16
|
| 51 |
+
},
|
| 52 |
+
"layer_norm_eps": 1e-12,
|
| 53 |
+
"max_position_embeddings": 512,
|
| 54 |
+
"model_type": "bert",
|
| 55 |
+
"num_attention_heads": 12,
|
| 56 |
+
"num_hidden_layers": 12,
|
| 57 |
+
"pad_token_id": 0,
|
| 58 |
+
"pooler_fc_size": 768,
|
| 59 |
+
"pooler_num_attention_heads": 12,
|
| 60 |
+
"pooler_num_fc_layers": 3,
|
| 61 |
+
"pooler_size_per_head": 128,
|
| 62 |
+
"pooler_type": "first_token_transform",
|
| 63 |
+
"position_embedding_type": "absolute",
|
| 64 |
+
"transformers_version": "4.56.0",
|
| 65 |
+
"type_vocab_size": 2,
|
| 66 |
+
"use_cache": false,
|
| 67 |
+
"vocab_size": 119547
|
| 68 |
+
}
|
financial-pii-ready/TokenBased-BERT/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 711491160
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:395beeb36e0647b8e6fee400c6be48d22264f6db457425982da6db1276ad1e90
|
| 3 |
size 711491160
|
financial-pii-ready/TokenBased-BERT/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5777
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f009f48664dfb0862500a88ed70a7098335f13a0191a4e36f095356e323cf542
|
| 3 |
size 5777
|