Upload folder using huggingface_hub
Browse files- financial-pii-ready/TokenBased-BERT/checkpoint-1407/config.json +72 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-1407/model.safetensors +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-1407/optimizer.pt +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-1407/rng_state.pth +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-1407/scheduler.pt +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-1407/tokenizer.json +0 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-1407/tokenizer_config.json +15 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-1407/trainer_state.json +56 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-1407/training_args.bin +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-2814/config.json +72 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-2814/model.safetensors +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-2814/optimizer.pt +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-2814/rng_state.pth +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-2814/scheduler.pt +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-2814/tokenizer.json +0 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-2814/tokenizer_config.json +15 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-2814/trainer_state.json +85 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-2814/training_args.bin +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-4221/config.json +72 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-4221/model.safetensors +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-4221/optimizer.pt +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-4221/rng_state.pth +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-4221/scheduler.pt +3 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-4221/tokenizer.json +0 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-4221/tokenizer_config.json +15 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-4221/trainer_state.json +114 -0
- financial-pii-ready/TokenBased-BERT/checkpoint-4221/training_args.bin +3 -0
- financial-pii-ready/TokenBased-BERT/config.json +72 -0
- financial-pii-ready/TokenBased-BERT/model.safetensors +3 -0
- financial-pii-ready/TokenBased-BERT/tokenizer.json +0 -0
- financial-pii-ready/TokenBased-BERT/tokenizer_config.json +15 -0
- financial-pii-ready/TokenBased-BERT/training_args.bin +3 -0
financial-pii-ready/TokenBased-BERT/checkpoint-1407/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"directionality": "bidi",
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_dropout_prob": 0.1,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"id2label": {
|
| 16 |
+
"0": "O",
|
| 17 |
+
"1": "B-ADDRESS",
|
| 18 |
+
"2": "I-ADDRESS",
|
| 19 |
+
"3": "B-COMPANY",
|
| 20 |
+
"4": "I-COMPANY",
|
| 21 |
+
"5": "B-CREDIT_CARD",
|
| 22 |
+
"6": "I-CREDIT_CARD",
|
| 23 |
+
"7": "B-EMAIL",
|
| 24 |
+
"8": "I-EMAIL",
|
| 25 |
+
"9": "B-NAME",
|
| 26 |
+
"10": "I-NAME",
|
| 27 |
+
"11": "B-PHONE",
|
| 28 |
+
"12": "I-PHONE",
|
| 29 |
+
"13": "B-SSN",
|
| 30 |
+
"14": "I-SSN",
|
| 31 |
+
"15": "B-URL",
|
| 32 |
+
"16": "I-URL"
|
| 33 |
+
},
|
| 34 |
+
"initializer_range": 0.02,
|
| 35 |
+
"intermediate_size": 3072,
|
| 36 |
+
"is_decoder": false,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"B-ADDRESS": 1,
|
| 39 |
+
"B-COMPANY": 3,
|
| 40 |
+
"B-CREDIT_CARD": 5,
|
| 41 |
+
"B-EMAIL": 7,
|
| 42 |
+
"B-NAME": 9,
|
| 43 |
+
"B-PHONE": 11,
|
| 44 |
+
"B-SSN": 13,
|
| 45 |
+
"B-URL": 15,
|
| 46 |
+
"I-ADDRESS": 2,
|
| 47 |
+
"I-COMPANY": 4,
|
| 48 |
+
"I-CREDIT_CARD": 6,
|
| 49 |
+
"I-EMAIL": 8,
|
| 50 |
+
"I-NAME": 10,
|
| 51 |
+
"I-PHONE": 12,
|
| 52 |
+
"I-SSN": 14,
|
| 53 |
+
"I-URL": 16,
|
| 54 |
+
"O": 0
|
| 55 |
+
},
|
| 56 |
+
"layer_norm_eps": 1e-12,
|
| 57 |
+
"max_position_embeddings": 512,
|
| 58 |
+
"model_type": "bert",
|
| 59 |
+
"num_attention_heads": 12,
|
| 60 |
+
"num_hidden_layers": 12,
|
| 61 |
+
"pad_token_id": 0,
|
| 62 |
+
"pooler_fc_size": 768,
|
| 63 |
+
"pooler_num_attention_heads": 12,
|
| 64 |
+
"pooler_num_fc_layers": 3,
|
| 65 |
+
"pooler_size_per_head": 128,
|
| 66 |
+
"pooler_type": "first_token_transform",
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.3.0",
|
| 69 |
+
"type_vocab_size": 2,
|
| 70 |
+
"use_cache": false,
|
| 71 |
+
"vocab_size": 119547
|
| 72 |
+
}
|
financial-pii-ready/TokenBased-BERT/checkpoint-1407/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:637645f07967566b9e1f9a788156a79bae317aeaa6d46850eb173675a41fe7bc
|
| 3 |
+
size 709127020
|
financial-pii-ready/TokenBased-BERT/checkpoint-1407/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ced38494c661e626ef98e4fcb880ffca426c4af53e952f8b0affc798c9cb3c32
|
| 3 |
+
size 1418377099
|
financial-pii-ready/TokenBased-BERT/checkpoint-1407/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2e9e20ad330a94fdb26e397df0e668f3424f8f42daa025d8535770293331d7d
|
| 3 |
+
size 14645
|
financial-pii-ready/TokenBased-BERT/checkpoint-1407/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6715a5a1995d6ff0489ea646102a508f7923079efd09ce587ae600a7314d1eb
|
| 3 |
+
size 1465
|
financial-pii-ready/TokenBased-BERT/checkpoint-1407/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready/TokenBased-BERT/checkpoint-1407/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"do_lower_case": false,
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
financial-pii-ready/TokenBased-BERT/checkpoint-1407/trainer_state.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1407,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.35536602700781805,
|
| 14 |
+
"grad_norm": 0.011881188489496708,
|
| 15 |
+
"learning_rate": 1.7635631366974652e-05,
|
| 16 |
+
"loss": 0.05762794876098633,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7107320540156361,
|
| 21 |
+
"grad_norm": 0.038132548332214355,
|
| 22 |
+
"learning_rate": 1.5266524520255865e-05,
|
| 23 |
+
"loss": 0.0012734627723693848,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_loss": 0.0003071273094974458,
|
| 29 |
+
"eval_runtime": 0.1309,
|
| 30 |
+
"eval_samples_per_second": 229.145,
|
| 31 |
+
"eval_steps_per_second": 30.553,
|
| 32 |
+
"step": 1407
|
| 33 |
+
}
|
| 34 |
+
],
|
| 35 |
+
"logging_steps": 500,
|
| 36 |
+
"max_steps": 4221,
|
| 37 |
+
"num_input_tokens_seen": 0,
|
| 38 |
+
"num_train_epochs": 3,
|
| 39 |
+
"save_steps": 500,
|
| 40 |
+
"stateful_callbacks": {
|
| 41 |
+
"TrainerControl": {
|
| 42 |
+
"args": {
|
| 43 |
+
"should_epoch_stop": false,
|
| 44 |
+
"should_evaluate": false,
|
| 45 |
+
"should_log": false,
|
| 46 |
+
"should_save": true,
|
| 47 |
+
"should_training_stop": false
|
| 48 |
+
},
|
| 49 |
+
"attributes": {}
|
| 50 |
+
}
|
| 51 |
+
},
|
| 52 |
+
"total_flos": 1.175994865152e+16,
|
| 53 |
+
"train_batch_size": 32,
|
| 54 |
+
"trial_name": null,
|
| 55 |
+
"trial_params": null
|
| 56 |
+
}
|
financial-pii-ready/TokenBased-BERT/checkpoint-1407/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f08d8fc75762bc57f6d363ddd3ef1bb5a7a079c46ea16fc29c2c922c42346bd
|
| 3 |
+
size 5201
|
financial-pii-ready/TokenBased-BERT/checkpoint-2814/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"directionality": "bidi",
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_dropout_prob": 0.1,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"id2label": {
|
| 16 |
+
"0": "O",
|
| 17 |
+
"1": "B-ADDRESS",
|
| 18 |
+
"2": "I-ADDRESS",
|
| 19 |
+
"3": "B-COMPANY",
|
| 20 |
+
"4": "I-COMPANY",
|
| 21 |
+
"5": "B-CREDIT_CARD",
|
| 22 |
+
"6": "I-CREDIT_CARD",
|
| 23 |
+
"7": "B-EMAIL",
|
| 24 |
+
"8": "I-EMAIL",
|
| 25 |
+
"9": "B-NAME",
|
| 26 |
+
"10": "I-NAME",
|
| 27 |
+
"11": "B-PHONE",
|
| 28 |
+
"12": "I-PHONE",
|
| 29 |
+
"13": "B-SSN",
|
| 30 |
+
"14": "I-SSN",
|
| 31 |
+
"15": "B-URL",
|
| 32 |
+
"16": "I-URL"
|
| 33 |
+
},
|
| 34 |
+
"initializer_range": 0.02,
|
| 35 |
+
"intermediate_size": 3072,
|
| 36 |
+
"is_decoder": false,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"B-ADDRESS": 1,
|
| 39 |
+
"B-COMPANY": 3,
|
| 40 |
+
"B-CREDIT_CARD": 5,
|
| 41 |
+
"B-EMAIL": 7,
|
| 42 |
+
"B-NAME": 9,
|
| 43 |
+
"B-PHONE": 11,
|
| 44 |
+
"B-SSN": 13,
|
| 45 |
+
"B-URL": 15,
|
| 46 |
+
"I-ADDRESS": 2,
|
| 47 |
+
"I-COMPANY": 4,
|
| 48 |
+
"I-CREDIT_CARD": 6,
|
| 49 |
+
"I-EMAIL": 8,
|
| 50 |
+
"I-NAME": 10,
|
| 51 |
+
"I-PHONE": 12,
|
| 52 |
+
"I-SSN": 14,
|
| 53 |
+
"I-URL": 16,
|
| 54 |
+
"O": 0
|
| 55 |
+
},
|
| 56 |
+
"layer_norm_eps": 1e-12,
|
| 57 |
+
"max_position_embeddings": 512,
|
| 58 |
+
"model_type": "bert",
|
| 59 |
+
"num_attention_heads": 12,
|
| 60 |
+
"num_hidden_layers": 12,
|
| 61 |
+
"pad_token_id": 0,
|
| 62 |
+
"pooler_fc_size": 768,
|
| 63 |
+
"pooler_num_attention_heads": 12,
|
| 64 |
+
"pooler_num_fc_layers": 3,
|
| 65 |
+
"pooler_size_per_head": 128,
|
| 66 |
+
"pooler_type": "first_token_transform",
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.3.0",
|
| 69 |
+
"type_vocab_size": 2,
|
| 70 |
+
"use_cache": false,
|
| 71 |
+
"vocab_size": 119547
|
| 72 |
+
}
|
financial-pii-ready/TokenBased-BERT/checkpoint-2814/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:23b8fd9f8303929872830879b3e1422a1f0a92887d52f946bde1db90d3920a65
|
| 3 |
+
size 709127020
|
financial-pii-ready/TokenBased-BERT/checkpoint-2814/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e01c0feffc67a29550e147fa3fbe29dfa30107f84596f9b33b4ccb27c11c2eaf
|
| 3 |
+
size 1418377099
|
financial-pii-ready/TokenBased-BERT/checkpoint-2814/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da3dd803c288f3bdbf75027cdb7d4043513e08ba70841481ca70c81158df4da9
|
| 3 |
+
size 14645
|
financial-pii-ready/TokenBased-BERT/checkpoint-2814/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14092e99b0f802054b8990cd9c06f451e94a8d55bcfbb236456db40050b79d23
|
| 3 |
+
size 1465
|
financial-pii-ready/TokenBased-BERT/checkpoint-2814/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready/TokenBased-BERT/checkpoint-2814/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"do_lower_case": false,
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
financial-pii-ready/TokenBased-BERT/checkpoint-2814/trainer_state.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2814,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.35536602700781805,
|
| 14 |
+
"grad_norm": 0.011881188489496708,
|
| 15 |
+
"learning_rate": 1.7635631366974652e-05,
|
| 16 |
+
"loss": 0.05762794876098633,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7107320540156361,
|
| 21 |
+
"grad_norm": 0.038132548332214355,
|
| 22 |
+
"learning_rate": 1.5266524520255865e-05,
|
| 23 |
+
"loss": 0.0012734627723693848,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_loss": 0.0003071273094974458,
|
| 29 |
+
"eval_runtime": 0.1309,
|
| 30 |
+
"eval_samples_per_second": 229.145,
|
| 31 |
+
"eval_steps_per_second": 30.553,
|
| 32 |
+
"step": 1407
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.0660980810234542,
|
| 36 |
+
"grad_norm": 0.002503013936802745,
|
| 37 |
+
"learning_rate": 1.2897417673537079e-05,
|
| 38 |
+
"loss": 0.0006923825144767761,
|
| 39 |
+
"step": 1500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.4214641080312722,
|
| 43 |
+
"grad_norm": 0.0017079897224903107,
|
| 44 |
+
"learning_rate": 1.052831082681829e-05,
|
| 45 |
+
"loss": 0.0004482545256614685,
|
| 46 |
+
"step": 2000
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.7768301350390903,
|
| 50 |
+
"grad_norm": 0.024019349366426468,
|
| 51 |
+
"learning_rate": 8.159203980099502e-06,
|
| 52 |
+
"loss": 0.0003331582248210907,
|
| 53 |
+
"step": 2500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 2.0,
|
| 57 |
+
"eval_loss": 0.00015163370699156076,
|
| 58 |
+
"eval_runtime": 0.1402,
|
| 59 |
+
"eval_samples_per_second": 214.008,
|
| 60 |
+
"eval_steps_per_second": 28.534,
|
| 61 |
+
"step": 2814
|
| 62 |
+
}
|
| 63 |
+
],
|
| 64 |
+
"logging_steps": 500,
|
| 65 |
+
"max_steps": 4221,
|
| 66 |
+
"num_input_tokens_seen": 0,
|
| 67 |
+
"num_train_epochs": 3,
|
| 68 |
+
"save_steps": 500,
|
| 69 |
+
"stateful_callbacks": {
|
| 70 |
+
"TrainerControl": {
|
| 71 |
+
"args": {
|
| 72 |
+
"should_epoch_stop": false,
|
| 73 |
+
"should_evaluate": false,
|
| 74 |
+
"should_log": false,
|
| 75 |
+
"should_save": true,
|
| 76 |
+
"should_training_stop": false
|
| 77 |
+
},
|
| 78 |
+
"attributes": {}
|
| 79 |
+
}
|
| 80 |
+
},
|
| 81 |
+
"total_flos": 2.351989730304e+16,
|
| 82 |
+
"train_batch_size": 32,
|
| 83 |
+
"trial_name": null,
|
| 84 |
+
"trial_params": null
|
| 85 |
+
}
|
financial-pii-ready/TokenBased-BERT/checkpoint-2814/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f08d8fc75762bc57f6d363ddd3ef1bb5a7a079c46ea16fc29c2c922c42346bd
|
| 3 |
+
size 5201
|
financial-pii-ready/TokenBased-BERT/checkpoint-4221/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"directionality": "bidi",
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_dropout_prob": 0.1,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"id2label": {
|
| 16 |
+
"0": "O",
|
| 17 |
+
"1": "B-ADDRESS",
|
| 18 |
+
"2": "I-ADDRESS",
|
| 19 |
+
"3": "B-COMPANY",
|
| 20 |
+
"4": "I-COMPANY",
|
| 21 |
+
"5": "B-CREDIT_CARD",
|
| 22 |
+
"6": "I-CREDIT_CARD",
|
| 23 |
+
"7": "B-EMAIL",
|
| 24 |
+
"8": "I-EMAIL",
|
| 25 |
+
"9": "B-NAME",
|
| 26 |
+
"10": "I-NAME",
|
| 27 |
+
"11": "B-PHONE",
|
| 28 |
+
"12": "I-PHONE",
|
| 29 |
+
"13": "B-SSN",
|
| 30 |
+
"14": "I-SSN",
|
| 31 |
+
"15": "B-URL",
|
| 32 |
+
"16": "I-URL"
|
| 33 |
+
},
|
| 34 |
+
"initializer_range": 0.02,
|
| 35 |
+
"intermediate_size": 3072,
|
| 36 |
+
"is_decoder": false,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"B-ADDRESS": 1,
|
| 39 |
+
"B-COMPANY": 3,
|
| 40 |
+
"B-CREDIT_CARD": 5,
|
| 41 |
+
"B-EMAIL": 7,
|
| 42 |
+
"B-NAME": 9,
|
| 43 |
+
"B-PHONE": 11,
|
| 44 |
+
"B-SSN": 13,
|
| 45 |
+
"B-URL": 15,
|
| 46 |
+
"I-ADDRESS": 2,
|
| 47 |
+
"I-COMPANY": 4,
|
| 48 |
+
"I-CREDIT_CARD": 6,
|
| 49 |
+
"I-EMAIL": 8,
|
| 50 |
+
"I-NAME": 10,
|
| 51 |
+
"I-PHONE": 12,
|
| 52 |
+
"I-SSN": 14,
|
| 53 |
+
"I-URL": 16,
|
| 54 |
+
"O": 0
|
| 55 |
+
},
|
| 56 |
+
"layer_norm_eps": 1e-12,
|
| 57 |
+
"max_position_embeddings": 512,
|
| 58 |
+
"model_type": "bert",
|
| 59 |
+
"num_attention_heads": 12,
|
| 60 |
+
"num_hidden_layers": 12,
|
| 61 |
+
"pad_token_id": 0,
|
| 62 |
+
"pooler_fc_size": 768,
|
| 63 |
+
"pooler_num_attention_heads": 12,
|
| 64 |
+
"pooler_num_fc_layers": 3,
|
| 65 |
+
"pooler_size_per_head": 128,
|
| 66 |
+
"pooler_type": "first_token_transform",
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.3.0",
|
| 69 |
+
"type_vocab_size": 2,
|
| 70 |
+
"use_cache": false,
|
| 71 |
+
"vocab_size": 119547
|
| 72 |
+
}
|
financial-pii-ready/TokenBased-BERT/checkpoint-4221/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca3cf7b3a56403c2fd93cd453b1f15f870b08c2e188395c9b1e789a977591dcc
|
| 3 |
+
size 709127020
|
financial-pii-ready/TokenBased-BERT/checkpoint-4221/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f82b81135bd2cbbdc2f08c5f666bcca8679418eda89bcb3beba197dd1aaf066c
|
| 3 |
+
size 1418377099
|
financial-pii-ready/TokenBased-BERT/checkpoint-4221/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a1438f6dec9837f6f095171cd868f490ca5a3fa918cf7c3575ac288b817be26
|
| 3 |
+
size 14645
|
financial-pii-ready/TokenBased-BERT/checkpoint-4221/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69b52d799f27b7647d5ff11de1e48fc331b2d5d2a0dcf8482a15b15f3fd4ef4d
|
| 3 |
+
size 1465
|
financial-pii-ready/TokenBased-BERT/checkpoint-4221/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready/TokenBased-BERT/checkpoint-4221/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"do_lower_case": false,
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
financial-pii-ready/TokenBased-BERT/checkpoint-4221/trainer_state.json
ADDED
|
@@ -0,0 +1,114 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 4221,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.35536602700781805,
|
| 14 |
+
"grad_norm": 0.011881188489496708,
|
| 15 |
+
"learning_rate": 1.7635631366974652e-05,
|
| 16 |
+
"loss": 0.05762794876098633,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7107320540156361,
|
| 21 |
+
"grad_norm": 0.038132548332214355,
|
| 22 |
+
"learning_rate": 1.5266524520255865e-05,
|
| 23 |
+
"loss": 0.0012734627723693848,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_loss": 0.0003071273094974458,
|
| 29 |
+
"eval_runtime": 0.1309,
|
| 30 |
+
"eval_samples_per_second": 229.145,
|
| 31 |
+
"eval_steps_per_second": 30.553,
|
| 32 |
+
"step": 1407
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.0660980810234542,
|
| 36 |
+
"grad_norm": 0.002503013936802745,
|
| 37 |
+
"learning_rate": 1.2897417673537079e-05,
|
| 38 |
+
"loss": 0.0006923825144767761,
|
| 39 |
+
"step": 1500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.4214641080312722,
|
| 43 |
+
"grad_norm": 0.0017079897224903107,
|
| 44 |
+
"learning_rate": 1.052831082681829e-05,
|
| 45 |
+
"loss": 0.0004482545256614685,
|
| 46 |
+
"step": 2000
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.7768301350390903,
|
| 50 |
+
"grad_norm": 0.024019349366426468,
|
| 51 |
+
"learning_rate": 8.159203980099502e-06,
|
| 52 |
+
"loss": 0.0003331582248210907,
|
| 53 |
+
"step": 2500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 2.0,
|
| 57 |
+
"eval_loss": 0.00015163370699156076,
|
| 58 |
+
"eval_runtime": 0.1402,
|
| 59 |
+
"eval_samples_per_second": 214.008,
|
| 60 |
+
"eval_steps_per_second": 28.534,
|
| 61 |
+
"step": 2814
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 2.1321961620469083,
|
| 65 |
+
"grad_norm": 0.006819023750722408,
|
| 66 |
+
"learning_rate": 5.790097133380715e-06,
|
| 67 |
+
"loss": 0.00026629400253295896,
|
| 68 |
+
"step": 3000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 2.487562189054726,
|
| 72 |
+
"grad_norm": 0.0235111266374588,
|
| 73 |
+
"learning_rate": 3.4209902866619288e-06,
|
| 74 |
+
"loss": 0.00020561683177947997,
|
| 75 |
+
"step": 3500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 2.8429282160625444,
|
| 79 |
+
"grad_norm": 0.05083432421088219,
|
| 80 |
+
"learning_rate": 1.0518834399431415e-06,
|
| 81 |
+
"loss": 0.00018332916498184205,
|
| 82 |
+
"step": 4000
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 3.0,
|
| 86 |
+
"eval_loss": 0.00039067131001502275,
|
| 87 |
+
"eval_runtime": 0.1864,
|
| 88 |
+
"eval_samples_per_second": 160.918,
|
| 89 |
+
"eval_steps_per_second": 21.456,
|
| 90 |
+
"step": 4221
|
| 91 |
+
}
|
| 92 |
+
],
|
| 93 |
+
"logging_steps": 500,
|
| 94 |
+
"max_steps": 4221,
|
| 95 |
+
"num_input_tokens_seen": 0,
|
| 96 |
+
"num_train_epochs": 3,
|
| 97 |
+
"save_steps": 500,
|
| 98 |
+
"stateful_callbacks": {
|
| 99 |
+
"TrainerControl": {
|
| 100 |
+
"args": {
|
| 101 |
+
"should_epoch_stop": false,
|
| 102 |
+
"should_evaluate": false,
|
| 103 |
+
"should_log": false,
|
| 104 |
+
"should_save": true,
|
| 105 |
+
"should_training_stop": true
|
| 106 |
+
},
|
| 107 |
+
"attributes": {}
|
| 108 |
+
}
|
| 109 |
+
},
|
| 110 |
+
"total_flos": 3.527984595456e+16,
|
| 111 |
+
"train_batch_size": 32,
|
| 112 |
+
"trial_name": null,
|
| 113 |
+
"trial_params": null
|
| 114 |
+
}
|
financial-pii-ready/TokenBased-BERT/checkpoint-4221/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f08d8fc75762bc57f6d363ddd3ef1bb5a7a079c46ea16fc29c2c922c42346bd
|
| 3 |
+
size 5201
|
financial-pii-ready/TokenBased-BERT/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"directionality": "bidi",
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_dropout_prob": 0.1,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"id2label": {
|
| 16 |
+
"0": "O",
|
| 17 |
+
"1": "B-ADDRESS",
|
| 18 |
+
"2": "I-ADDRESS",
|
| 19 |
+
"3": "B-COMPANY",
|
| 20 |
+
"4": "I-COMPANY",
|
| 21 |
+
"5": "B-CREDIT_CARD",
|
| 22 |
+
"6": "I-CREDIT_CARD",
|
| 23 |
+
"7": "B-EMAIL",
|
| 24 |
+
"8": "I-EMAIL",
|
| 25 |
+
"9": "B-NAME",
|
| 26 |
+
"10": "I-NAME",
|
| 27 |
+
"11": "B-PHONE",
|
| 28 |
+
"12": "I-PHONE",
|
| 29 |
+
"13": "B-SSN",
|
| 30 |
+
"14": "I-SSN",
|
| 31 |
+
"15": "B-URL",
|
| 32 |
+
"16": "I-URL"
|
| 33 |
+
},
|
| 34 |
+
"initializer_range": 0.02,
|
| 35 |
+
"intermediate_size": 3072,
|
| 36 |
+
"is_decoder": false,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"B-ADDRESS": 1,
|
| 39 |
+
"B-COMPANY": 3,
|
| 40 |
+
"B-CREDIT_CARD": 5,
|
| 41 |
+
"B-EMAIL": 7,
|
| 42 |
+
"B-NAME": 9,
|
| 43 |
+
"B-PHONE": 11,
|
| 44 |
+
"B-SSN": 13,
|
| 45 |
+
"B-URL": 15,
|
| 46 |
+
"I-ADDRESS": 2,
|
| 47 |
+
"I-COMPANY": 4,
|
| 48 |
+
"I-CREDIT_CARD": 6,
|
| 49 |
+
"I-EMAIL": 8,
|
| 50 |
+
"I-NAME": 10,
|
| 51 |
+
"I-PHONE": 12,
|
| 52 |
+
"I-SSN": 14,
|
| 53 |
+
"I-URL": 16,
|
| 54 |
+
"O": 0
|
| 55 |
+
},
|
| 56 |
+
"layer_norm_eps": 1e-12,
|
| 57 |
+
"max_position_embeddings": 512,
|
| 58 |
+
"model_type": "bert",
|
| 59 |
+
"num_attention_heads": 12,
|
| 60 |
+
"num_hidden_layers": 12,
|
| 61 |
+
"pad_token_id": 0,
|
| 62 |
+
"pooler_fc_size": 768,
|
| 63 |
+
"pooler_num_attention_heads": 12,
|
| 64 |
+
"pooler_num_fc_layers": 3,
|
| 65 |
+
"pooler_size_per_head": 128,
|
| 66 |
+
"pooler_type": "first_token_transform",
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.3.0",
|
| 69 |
+
"type_vocab_size": 2,
|
| 70 |
+
"use_cache": false,
|
| 71 |
+
"vocab_size": 119547
|
| 72 |
+
}
|
financial-pii-ready/TokenBased-BERT/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca3cf7b3a56403c2fd93cd453b1f15f870b08c2e188395c9b1e789a977591dcc
|
| 3 |
+
size 709127020
|
financial-pii-ready/TokenBased-BERT/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready/TokenBased-BERT/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"do_lower_case": false,
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
financial-pii-ready/TokenBased-BERT/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f08d8fc75762bc57f6d363ddd3ef1bb5a7a079c46ea16fc29c2c922c42346bd
|
| 3 |
+
size 5201
|