Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/config.json +129 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/rng_state.pth +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json +0 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer_config.json +23 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/trainer_state.json +48 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/config.json +129 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/model.safetensors +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/optimizer.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/rng_state.pth +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/scheduler.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/tokenizer.json +0 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/tokenizer_config.json +23 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/trainer_state.json +55 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/training_args.bin +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/config.json +129 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/model.safetensors +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/optimizer.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/rng_state.pth +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/scheduler.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/tokenizer.json +0 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/tokenizer_config.json +23 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/trainer_state.json +62 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/training_args.bin +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/config.json +129 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/model.safetensors +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/optimizer.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/rng_state.pth +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/scheduler.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/tokenizer.json +0 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/tokenizer_config.json +23 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/trainer_state.json +69 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/training_args.bin +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/config.json +129 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/model.safetensors +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/optimizer.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/rng_state.pth +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/scheduler.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/tokenizer.json +0 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/tokenizer_config.json +23 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/trainer_state.json +76 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/training_args.bin +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/config.json +129 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/model.safetensors +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/optimizer.pt +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/rng_state.pth +3 -0
- gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/scheduler.pt +3 -0
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ACCOUNT_NUMBER",
|
| 15 |
+
"2": "ADDRESS",
|
| 16 |
+
"3": "API_KEY",
|
| 17 |
+
"4": "BANK_ROUTING_NUMBER",
|
| 18 |
+
"5": "BIOMETRIC_IDENTIFIER",
|
| 19 |
+
"6": "CERTIFICATE_LICENSE_NUMBER",
|
| 20 |
+
"7": "CITY",
|
| 21 |
+
"8": "COMPANY_NAME",
|
| 22 |
+
"9": "COORDINATE",
|
| 23 |
+
"10": "COUNTRY",
|
| 24 |
+
"11": "CREDIT_CARD_NUMBER",
|
| 25 |
+
"12": "CUSTOMER_ID",
|
| 26 |
+
"13": "CVV",
|
| 27 |
+
"14": "DATE",
|
| 28 |
+
"15": "DATE_OF_BIRTH",
|
| 29 |
+
"16": "DATE_TIME",
|
| 30 |
+
"17": "DEVICE_IDENTIFIER",
|
| 31 |
+
"18": "EMAIL",
|
| 32 |
+
"19": "EMPLOYEE_ID",
|
| 33 |
+
"20": "FIRST_NAME",
|
| 34 |
+
"21": "HEALTH_PLAN_BENEFICIARY_NUMBER",
|
| 35 |
+
"22": "IPV4",
|
| 36 |
+
"23": "IPV6",
|
| 37 |
+
"24": "LAST_NAME",
|
| 38 |
+
"25": "LICENSE_PLATE",
|
| 39 |
+
"26": "MEDICAL_RECORD_NUMBER",
|
| 40 |
+
"27": "NAME",
|
| 41 |
+
"28": "NATIONAL_ID",
|
| 42 |
+
"29": "PASSWORD",
|
| 43 |
+
"30": "PHONE_NUMBER",
|
| 44 |
+
"31": "PIN",
|
| 45 |
+
"32": "POSTCODE",
|
| 46 |
+
"33": "SSN",
|
| 47 |
+
"34": "STATE",
|
| 48 |
+
"35": "STREET_ADDRESS",
|
| 49 |
+
"36": "SWIFT_BIC",
|
| 50 |
+
"37": "TAX_ID",
|
| 51 |
+
"38": "TIME",
|
| 52 |
+
"39": "UNIQUE_IDENTIFIER",
|
| 53 |
+
"40": "URL",
|
| 54 |
+
"41": "USER_NAME",
|
| 55 |
+
"42": "VEHICLE_IDENTIFIER"
|
| 56 |
+
},
|
| 57 |
+
"initializer_range": 0.02,
|
| 58 |
+
"intermediate_size": 3072,
|
| 59 |
+
"label2id": {
|
| 60 |
+
"ACCOUNT_NUMBER": 1,
|
| 61 |
+
"ADDRESS": 2,
|
| 62 |
+
"API_KEY": 3,
|
| 63 |
+
"BANK_ROUTING_NUMBER": 4,
|
| 64 |
+
"BIOMETRIC_IDENTIFIER": 5,
|
| 65 |
+
"CERTIFICATE_LICENSE_NUMBER": 6,
|
| 66 |
+
"CITY": 7,
|
| 67 |
+
"COMPANY_NAME": 8,
|
| 68 |
+
"COORDINATE": 9,
|
| 69 |
+
"COUNTRY": 10,
|
| 70 |
+
"CREDIT_CARD_NUMBER": 11,
|
| 71 |
+
"CUSTOMER_ID": 12,
|
| 72 |
+
"CVV": 13,
|
| 73 |
+
"DATE": 14,
|
| 74 |
+
"DATE_OF_BIRTH": 15,
|
| 75 |
+
"DATE_TIME": 16,
|
| 76 |
+
"DEVICE_IDENTIFIER": 17,
|
| 77 |
+
"EMAIL": 18,
|
| 78 |
+
"EMPLOYEE_ID": 19,
|
| 79 |
+
"FIRST_NAME": 20,
|
| 80 |
+
"HEALTH_PLAN_BENEFICIARY_NUMBER": 21,
|
| 81 |
+
"IPV4": 22,
|
| 82 |
+
"IPV6": 23,
|
| 83 |
+
"LAST_NAME": 24,
|
| 84 |
+
"LICENSE_PLATE": 25,
|
| 85 |
+
"MEDICAL_RECORD_NUMBER": 26,
|
| 86 |
+
"NAME": 27,
|
| 87 |
+
"NATIONAL_ID": 28,
|
| 88 |
+
"O": 0,
|
| 89 |
+
"PASSWORD": 29,
|
| 90 |
+
"PHONE_NUMBER": 30,
|
| 91 |
+
"PIN": 31,
|
| 92 |
+
"POSTCODE": 32,
|
| 93 |
+
"SSN": 33,
|
| 94 |
+
"STATE": 34,
|
| 95 |
+
"STREET_ADDRESS": 35,
|
| 96 |
+
"SWIFT_BIC": 36,
|
| 97 |
+
"TAX_ID": 37,
|
| 98 |
+
"TIME": 38,
|
| 99 |
+
"UNIQUE_IDENTIFIER": 39,
|
| 100 |
+
"URL": 40,
|
| 101 |
+
"USER_NAME": 41,
|
| 102 |
+
"VEHICLE_IDENTIFIER": 42
|
| 103 |
+
},
|
| 104 |
+
"layer_norm_eps": 1e-07,
|
| 105 |
+
"legacy": true,
|
| 106 |
+
"max_position_embeddings": 512,
|
| 107 |
+
"max_relative_positions": -1,
|
| 108 |
+
"model_type": "deberta-v2",
|
| 109 |
+
"norm_rel_ebd": "layer_norm",
|
| 110 |
+
"num_attention_heads": 12,
|
| 111 |
+
"num_hidden_layers": 12,
|
| 112 |
+
"pad_token_id": 0,
|
| 113 |
+
"pooler_dropout": 0,
|
| 114 |
+
"pooler_hidden_act": "gelu",
|
| 115 |
+
"pooler_hidden_size": 768,
|
| 116 |
+
"pos_att_type": [
|
| 117 |
+
"p2c",
|
| 118 |
+
"c2p"
|
| 119 |
+
],
|
| 120 |
+
"position_biased_input": false,
|
| 121 |
+
"position_buckets": 256,
|
| 122 |
+
"relative_attention": true,
|
| 123 |
+
"share_att_key": true,
|
| 124 |
+
"tie_word_embeddings": true,
|
| 125 |
+
"transformers_version": "5.3.0",
|
| 126 |
+
"type_vocab_size": 0,
|
| 127 |
+
"use_cache": false,
|
| 128 |
+
"vocab_size": 128100
|
| 129 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95c5d29c33fdeb4b9612dde5a8b89a7edec751ce7f706778b4e0f7e943ac2b92
|
| 3 |
+
size 741461252
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1a5dbef385605180de06af888e928d2fb394b5d89f0615c6f7dc9e5f0f11449c
|
| 3 |
+
size 1483046155
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:97fd2fd7e23183b8e6a6f532d64bd1d242ceb3984f6b1ad076e4d19c204207e9
|
| 3 |
+
size 14645
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6eaea66deaece4e5fc233aefeecc3d74f8e5cba57072d8645150e28e7a7058e2
|
| 3 |
+
size 1465
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "[CLS]",
|
| 5 |
+
"cls_token": "[CLS]",
|
| 6 |
+
"do_lower_case": false,
|
| 7 |
+
"eos_token": "[SEP]",
|
| 8 |
+
"extra_special_tokens": [
|
| 9 |
+
"[PAD]",
|
| 10 |
+
"[CLS]",
|
| 11 |
+
"[SEP]"
|
| 12 |
+
],
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"mask_token": "[MASK]",
|
| 15 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 16 |
+
"pad_token": "[PAD]",
|
| 17 |
+
"sep_token": "[SEP]",
|
| 18 |
+
"split_by_punct": false,
|
| 19 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 20 |
+
"unk_id": 3,
|
| 21 |
+
"unk_token": "[UNK]",
|
| 22 |
+
"vocab_type": "spm"
|
| 23 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.35549235691432635,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.17774617845716317,
|
| 14 |
+
"grad_norm": 0.011032640002667904,
|
| 15 |
+
"learning_rate": 1.8226093138997513e-05,
|
| 16 |
+
"loss": 0.007762076377868653,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.35549235691432635,
|
| 21 |
+
"grad_norm": 0.012853773310780525,
|
| 22 |
+
"learning_rate": 1.6448631354425882e-05,
|
| 23 |
+
"loss": 0.0016952511072158814,
|
| 24 |
+
"step": 1000
|
| 25 |
+
}
|
| 26 |
+
],
|
| 27 |
+
"logging_steps": 500,
|
| 28 |
+
"max_steps": 5626,
|
| 29 |
+
"num_input_tokens_seen": 0,
|
| 30 |
+
"num_train_epochs": 2,
|
| 31 |
+
"save_steps": 500,
|
| 32 |
+
"stateful_callbacks": {
|
| 33 |
+
"TrainerControl": {
|
| 34 |
+
"args": {
|
| 35 |
+
"should_epoch_stop": false,
|
| 36 |
+
"should_evaluate": false,
|
| 37 |
+
"should_log": false,
|
| 38 |
+
"should_save": true,
|
| 39 |
+
"should_training_stop": false
|
| 40 |
+
},
|
| 41 |
+
"attributes": {}
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"total_flos": 3125930394782400.0,
|
| 45 |
+
"train_batch_size": 32,
|
| 46 |
+
"trial_name": null,
|
| 47 |
+
"trial_params": null
|
| 48 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a369d04c392a152038f999ecba936911bcde812334c182dac021a19c2d915389
|
| 3 |
+
size 5201
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ACCOUNT_NUMBER",
|
| 15 |
+
"2": "ADDRESS",
|
| 16 |
+
"3": "API_KEY",
|
| 17 |
+
"4": "BANK_ROUTING_NUMBER",
|
| 18 |
+
"5": "BIOMETRIC_IDENTIFIER",
|
| 19 |
+
"6": "CERTIFICATE_LICENSE_NUMBER",
|
| 20 |
+
"7": "CITY",
|
| 21 |
+
"8": "COMPANY_NAME",
|
| 22 |
+
"9": "COORDINATE",
|
| 23 |
+
"10": "COUNTRY",
|
| 24 |
+
"11": "CREDIT_CARD_NUMBER",
|
| 25 |
+
"12": "CUSTOMER_ID",
|
| 26 |
+
"13": "CVV",
|
| 27 |
+
"14": "DATE",
|
| 28 |
+
"15": "DATE_OF_BIRTH",
|
| 29 |
+
"16": "DATE_TIME",
|
| 30 |
+
"17": "DEVICE_IDENTIFIER",
|
| 31 |
+
"18": "EMAIL",
|
| 32 |
+
"19": "EMPLOYEE_ID",
|
| 33 |
+
"20": "FIRST_NAME",
|
| 34 |
+
"21": "HEALTH_PLAN_BENEFICIARY_NUMBER",
|
| 35 |
+
"22": "IPV4",
|
| 36 |
+
"23": "IPV6",
|
| 37 |
+
"24": "LAST_NAME",
|
| 38 |
+
"25": "LICENSE_PLATE",
|
| 39 |
+
"26": "MEDICAL_RECORD_NUMBER",
|
| 40 |
+
"27": "NAME",
|
| 41 |
+
"28": "NATIONAL_ID",
|
| 42 |
+
"29": "PASSWORD",
|
| 43 |
+
"30": "PHONE_NUMBER",
|
| 44 |
+
"31": "PIN",
|
| 45 |
+
"32": "POSTCODE",
|
| 46 |
+
"33": "SSN",
|
| 47 |
+
"34": "STATE",
|
| 48 |
+
"35": "STREET_ADDRESS",
|
| 49 |
+
"36": "SWIFT_BIC",
|
| 50 |
+
"37": "TAX_ID",
|
| 51 |
+
"38": "TIME",
|
| 52 |
+
"39": "UNIQUE_IDENTIFIER",
|
| 53 |
+
"40": "URL",
|
| 54 |
+
"41": "USER_NAME",
|
| 55 |
+
"42": "VEHICLE_IDENTIFIER"
|
| 56 |
+
},
|
| 57 |
+
"initializer_range": 0.02,
|
| 58 |
+
"intermediate_size": 3072,
|
| 59 |
+
"label2id": {
|
| 60 |
+
"ACCOUNT_NUMBER": 1,
|
| 61 |
+
"ADDRESS": 2,
|
| 62 |
+
"API_KEY": 3,
|
| 63 |
+
"BANK_ROUTING_NUMBER": 4,
|
| 64 |
+
"BIOMETRIC_IDENTIFIER": 5,
|
| 65 |
+
"CERTIFICATE_LICENSE_NUMBER": 6,
|
| 66 |
+
"CITY": 7,
|
| 67 |
+
"COMPANY_NAME": 8,
|
| 68 |
+
"COORDINATE": 9,
|
| 69 |
+
"COUNTRY": 10,
|
| 70 |
+
"CREDIT_CARD_NUMBER": 11,
|
| 71 |
+
"CUSTOMER_ID": 12,
|
| 72 |
+
"CVV": 13,
|
| 73 |
+
"DATE": 14,
|
| 74 |
+
"DATE_OF_BIRTH": 15,
|
| 75 |
+
"DATE_TIME": 16,
|
| 76 |
+
"DEVICE_IDENTIFIER": 17,
|
| 77 |
+
"EMAIL": 18,
|
| 78 |
+
"EMPLOYEE_ID": 19,
|
| 79 |
+
"FIRST_NAME": 20,
|
| 80 |
+
"HEALTH_PLAN_BENEFICIARY_NUMBER": 21,
|
| 81 |
+
"IPV4": 22,
|
| 82 |
+
"IPV6": 23,
|
| 83 |
+
"LAST_NAME": 24,
|
| 84 |
+
"LICENSE_PLATE": 25,
|
| 85 |
+
"MEDICAL_RECORD_NUMBER": 26,
|
| 86 |
+
"NAME": 27,
|
| 87 |
+
"NATIONAL_ID": 28,
|
| 88 |
+
"O": 0,
|
| 89 |
+
"PASSWORD": 29,
|
| 90 |
+
"PHONE_NUMBER": 30,
|
| 91 |
+
"PIN": 31,
|
| 92 |
+
"POSTCODE": 32,
|
| 93 |
+
"SSN": 33,
|
| 94 |
+
"STATE": 34,
|
| 95 |
+
"STREET_ADDRESS": 35,
|
| 96 |
+
"SWIFT_BIC": 36,
|
| 97 |
+
"TAX_ID": 37,
|
| 98 |
+
"TIME": 38,
|
| 99 |
+
"UNIQUE_IDENTIFIER": 39,
|
| 100 |
+
"URL": 40,
|
| 101 |
+
"USER_NAME": 41,
|
| 102 |
+
"VEHICLE_IDENTIFIER": 42
|
| 103 |
+
},
|
| 104 |
+
"layer_norm_eps": 1e-07,
|
| 105 |
+
"legacy": true,
|
| 106 |
+
"max_position_embeddings": 512,
|
| 107 |
+
"max_relative_positions": -1,
|
| 108 |
+
"model_type": "deberta-v2",
|
| 109 |
+
"norm_rel_ebd": "layer_norm",
|
| 110 |
+
"num_attention_heads": 12,
|
| 111 |
+
"num_hidden_layers": 12,
|
| 112 |
+
"pad_token_id": 0,
|
| 113 |
+
"pooler_dropout": 0,
|
| 114 |
+
"pooler_hidden_act": "gelu",
|
| 115 |
+
"pooler_hidden_size": 768,
|
| 116 |
+
"pos_att_type": [
|
| 117 |
+
"p2c",
|
| 118 |
+
"c2p"
|
| 119 |
+
],
|
| 120 |
+
"position_biased_input": false,
|
| 121 |
+
"position_buckets": 256,
|
| 122 |
+
"relative_attention": true,
|
| 123 |
+
"share_att_key": true,
|
| 124 |
+
"tie_word_embeddings": true,
|
| 125 |
+
"transformers_version": "5.3.0",
|
| 126 |
+
"type_vocab_size": 0,
|
| 127 |
+
"use_cache": false,
|
| 128 |
+
"vocab_size": 128100
|
| 129 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4cb6a21895cb18780a20a6551120fb9a5fde67cf1b5947bfb2fb76752cfebfbd
|
| 3 |
+
size 741461252
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:155a05068a05978ce5436b32233fe7596edcf26e8a4fdfbefdd1cf0cf4e21917
|
| 3 |
+
size 1483046155
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7aed804a648a35d09285c63b003968bbd59d32c51e826eb711d02a234520652b
|
| 3 |
+
size 14645
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c84fd6f3f0458af2210e88908f8b85a694d875e45b7e83cf4662e94ce0debdb
|
| 3 |
+
size 1465
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/tokenizer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "[CLS]",
|
| 5 |
+
"cls_token": "[CLS]",
|
| 6 |
+
"do_lower_case": false,
|
| 7 |
+
"eos_token": "[SEP]",
|
| 8 |
+
"extra_special_tokens": [
|
| 9 |
+
"[PAD]",
|
| 10 |
+
"[CLS]",
|
| 11 |
+
"[SEP]"
|
| 12 |
+
],
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"mask_token": "[MASK]",
|
| 15 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 16 |
+
"pad_token": "[PAD]",
|
| 17 |
+
"sep_token": "[SEP]",
|
| 18 |
+
"split_by_punct": false,
|
| 19 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 20 |
+
"unk_id": 3,
|
| 21 |
+
"unk_token": "[UNK]",
|
| 22 |
+
"vocab_type": "spm"
|
| 23 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/trainer_state.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.5332385353714895,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.17774617845716317,
|
| 14 |
+
"grad_norm": 0.011032640002667904,
|
| 15 |
+
"learning_rate": 1.8226093138997513e-05,
|
| 16 |
+
"loss": 0.007762076377868653,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.35549235691432635,
|
| 21 |
+
"grad_norm": 0.012853773310780525,
|
| 22 |
+
"learning_rate": 1.6448631354425882e-05,
|
| 23 |
+
"loss": 0.0016952511072158814,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5332385353714895,
|
| 28 |
+
"grad_norm": 0.00953384954482317,
|
| 29 |
+
"learning_rate": 1.4671169569854249e-05,
|
| 30 |
+
"loss": 0.0008515777587890625,
|
| 31 |
+
"step": 1500
|
| 32 |
+
}
|
| 33 |
+
],
|
| 34 |
+
"logging_steps": 500,
|
| 35 |
+
"max_steps": 5626,
|
| 36 |
+
"num_input_tokens_seen": 0,
|
| 37 |
+
"num_train_epochs": 2,
|
| 38 |
+
"save_steps": 500,
|
| 39 |
+
"stateful_callbacks": {
|
| 40 |
+
"TrainerControl": {
|
| 41 |
+
"args": {
|
| 42 |
+
"should_epoch_stop": false,
|
| 43 |
+
"should_evaluate": false,
|
| 44 |
+
"should_log": false,
|
| 45 |
+
"should_save": true,
|
| 46 |
+
"should_training_stop": false
|
| 47 |
+
},
|
| 48 |
+
"attributes": {}
|
| 49 |
+
}
|
| 50 |
+
},
|
| 51 |
+
"total_flos": 4661911952689728.0,
|
| 52 |
+
"train_batch_size": 32,
|
| 53 |
+
"trial_name": null,
|
| 54 |
+
"trial_params": null
|
| 55 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a369d04c392a152038f999ecba936911bcde812334c182dac021a19c2d915389
|
| 3 |
+
size 5201
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ACCOUNT_NUMBER",
|
| 15 |
+
"2": "ADDRESS",
|
| 16 |
+
"3": "API_KEY",
|
| 17 |
+
"4": "BANK_ROUTING_NUMBER",
|
| 18 |
+
"5": "BIOMETRIC_IDENTIFIER",
|
| 19 |
+
"6": "CERTIFICATE_LICENSE_NUMBER",
|
| 20 |
+
"7": "CITY",
|
| 21 |
+
"8": "COMPANY_NAME",
|
| 22 |
+
"9": "COORDINATE",
|
| 23 |
+
"10": "COUNTRY",
|
| 24 |
+
"11": "CREDIT_CARD_NUMBER",
|
| 25 |
+
"12": "CUSTOMER_ID",
|
| 26 |
+
"13": "CVV",
|
| 27 |
+
"14": "DATE",
|
| 28 |
+
"15": "DATE_OF_BIRTH",
|
| 29 |
+
"16": "DATE_TIME",
|
| 30 |
+
"17": "DEVICE_IDENTIFIER",
|
| 31 |
+
"18": "EMAIL",
|
| 32 |
+
"19": "EMPLOYEE_ID",
|
| 33 |
+
"20": "FIRST_NAME",
|
| 34 |
+
"21": "HEALTH_PLAN_BENEFICIARY_NUMBER",
|
| 35 |
+
"22": "IPV4",
|
| 36 |
+
"23": "IPV6",
|
| 37 |
+
"24": "LAST_NAME",
|
| 38 |
+
"25": "LICENSE_PLATE",
|
| 39 |
+
"26": "MEDICAL_RECORD_NUMBER",
|
| 40 |
+
"27": "NAME",
|
| 41 |
+
"28": "NATIONAL_ID",
|
| 42 |
+
"29": "PASSWORD",
|
| 43 |
+
"30": "PHONE_NUMBER",
|
| 44 |
+
"31": "PIN",
|
| 45 |
+
"32": "POSTCODE",
|
| 46 |
+
"33": "SSN",
|
| 47 |
+
"34": "STATE",
|
| 48 |
+
"35": "STREET_ADDRESS",
|
| 49 |
+
"36": "SWIFT_BIC",
|
| 50 |
+
"37": "TAX_ID",
|
| 51 |
+
"38": "TIME",
|
| 52 |
+
"39": "UNIQUE_IDENTIFIER",
|
| 53 |
+
"40": "URL",
|
| 54 |
+
"41": "USER_NAME",
|
| 55 |
+
"42": "VEHICLE_IDENTIFIER"
|
| 56 |
+
},
|
| 57 |
+
"initializer_range": 0.02,
|
| 58 |
+
"intermediate_size": 3072,
|
| 59 |
+
"label2id": {
|
| 60 |
+
"ACCOUNT_NUMBER": 1,
|
| 61 |
+
"ADDRESS": 2,
|
| 62 |
+
"API_KEY": 3,
|
| 63 |
+
"BANK_ROUTING_NUMBER": 4,
|
| 64 |
+
"BIOMETRIC_IDENTIFIER": 5,
|
| 65 |
+
"CERTIFICATE_LICENSE_NUMBER": 6,
|
| 66 |
+
"CITY": 7,
|
| 67 |
+
"COMPANY_NAME": 8,
|
| 68 |
+
"COORDINATE": 9,
|
| 69 |
+
"COUNTRY": 10,
|
| 70 |
+
"CREDIT_CARD_NUMBER": 11,
|
| 71 |
+
"CUSTOMER_ID": 12,
|
| 72 |
+
"CVV": 13,
|
| 73 |
+
"DATE": 14,
|
| 74 |
+
"DATE_OF_BIRTH": 15,
|
| 75 |
+
"DATE_TIME": 16,
|
| 76 |
+
"DEVICE_IDENTIFIER": 17,
|
| 77 |
+
"EMAIL": 18,
|
| 78 |
+
"EMPLOYEE_ID": 19,
|
| 79 |
+
"FIRST_NAME": 20,
|
| 80 |
+
"HEALTH_PLAN_BENEFICIARY_NUMBER": 21,
|
| 81 |
+
"IPV4": 22,
|
| 82 |
+
"IPV6": 23,
|
| 83 |
+
"LAST_NAME": 24,
|
| 84 |
+
"LICENSE_PLATE": 25,
|
| 85 |
+
"MEDICAL_RECORD_NUMBER": 26,
|
| 86 |
+
"NAME": 27,
|
| 87 |
+
"NATIONAL_ID": 28,
|
| 88 |
+
"O": 0,
|
| 89 |
+
"PASSWORD": 29,
|
| 90 |
+
"PHONE_NUMBER": 30,
|
| 91 |
+
"PIN": 31,
|
| 92 |
+
"POSTCODE": 32,
|
| 93 |
+
"SSN": 33,
|
| 94 |
+
"STATE": 34,
|
| 95 |
+
"STREET_ADDRESS": 35,
|
| 96 |
+
"SWIFT_BIC": 36,
|
| 97 |
+
"TAX_ID": 37,
|
| 98 |
+
"TIME": 38,
|
| 99 |
+
"UNIQUE_IDENTIFIER": 39,
|
| 100 |
+
"URL": 40,
|
| 101 |
+
"USER_NAME": 41,
|
| 102 |
+
"VEHICLE_IDENTIFIER": 42
|
| 103 |
+
},
|
| 104 |
+
"layer_norm_eps": 1e-07,
|
| 105 |
+
"legacy": true,
|
| 106 |
+
"max_position_embeddings": 512,
|
| 107 |
+
"max_relative_positions": -1,
|
| 108 |
+
"model_type": "deberta-v2",
|
| 109 |
+
"norm_rel_ebd": "layer_norm",
|
| 110 |
+
"num_attention_heads": 12,
|
| 111 |
+
"num_hidden_layers": 12,
|
| 112 |
+
"pad_token_id": 0,
|
| 113 |
+
"pooler_dropout": 0,
|
| 114 |
+
"pooler_hidden_act": "gelu",
|
| 115 |
+
"pooler_hidden_size": 768,
|
| 116 |
+
"pos_att_type": [
|
| 117 |
+
"p2c",
|
| 118 |
+
"c2p"
|
| 119 |
+
],
|
| 120 |
+
"position_biased_input": false,
|
| 121 |
+
"position_buckets": 256,
|
| 122 |
+
"relative_attention": true,
|
| 123 |
+
"share_att_key": true,
|
| 124 |
+
"tie_word_embeddings": true,
|
| 125 |
+
"transformers_version": "5.3.0",
|
| 126 |
+
"type_vocab_size": 0,
|
| 127 |
+
"use_cache": false,
|
| 128 |
+
"vocab_size": 128100
|
| 129 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90a5e540ce456767343d519c8b1c59efc753e1143a87b7cab153c0771c9dfba0
|
| 3 |
+
size 741461252
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:56ec6eae61d0fc70c6e138f3720a00e07c3d69ea69d580db77b7f5a1570d88bb
|
| 3 |
+
size 1483046155
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:452ce4eb1c812b9f74c4b09c203436cb1ab1f373bbce5c52986383f6aecdbcdc
|
| 3 |
+
size 14645
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:035e3e19fd9659817d0e37975e42543bfb350b6f3af3c098dcc7ca622b2e9b4e
|
| 3 |
+
size 1465
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "[CLS]",
|
| 5 |
+
"cls_token": "[CLS]",
|
| 6 |
+
"do_lower_case": false,
|
| 7 |
+
"eos_token": "[SEP]",
|
| 8 |
+
"extra_special_tokens": [
|
| 9 |
+
"[PAD]",
|
| 10 |
+
"[CLS]",
|
| 11 |
+
"[SEP]"
|
| 12 |
+
],
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"mask_token": "[MASK]",
|
| 15 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 16 |
+
"pad_token": "[PAD]",
|
| 17 |
+
"sep_token": "[SEP]",
|
| 18 |
+
"split_by_punct": false,
|
| 19 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 20 |
+
"unk_id": 3,
|
| 21 |
+
"unk_token": "[UNK]",
|
| 22 |
+
"vocab_type": "spm"
|
| 23 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/trainer_state.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.7109847138286527,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.17774617845716317,
|
| 14 |
+
"grad_norm": 0.011032640002667904,
|
| 15 |
+
"learning_rate": 1.8226093138997513e-05,
|
| 16 |
+
"loss": 0.007762076377868653,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.35549235691432635,
|
| 21 |
+
"grad_norm": 0.012853773310780525,
|
| 22 |
+
"learning_rate": 1.6448631354425882e-05,
|
| 23 |
+
"loss": 0.0016952511072158814,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5332385353714895,
|
| 28 |
+
"grad_norm": 0.00953384954482317,
|
| 29 |
+
"learning_rate": 1.4671169569854249e-05,
|
| 30 |
+
"loss": 0.0008515777587890625,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.7109847138286527,
|
| 35 |
+
"grad_norm": 0.012823447585105896,
|
| 36 |
+
"learning_rate": 1.2893707785282616e-05,
|
| 37 |
+
"loss": 0.0006071038842201233,
|
| 38 |
+
"step": 2000
|
| 39 |
+
}
|
| 40 |
+
],
|
| 41 |
+
"logging_steps": 500,
|
| 42 |
+
"max_steps": 5626,
|
| 43 |
+
"num_input_tokens_seen": 0,
|
| 44 |
+
"num_train_epochs": 2,
|
| 45 |
+
"save_steps": 500,
|
| 46 |
+
"stateful_callbacks": {
|
| 47 |
+
"TrainerControl": {
|
| 48 |
+
"args": {
|
| 49 |
+
"should_epoch_stop": false,
|
| 50 |
+
"should_evaluate": false,
|
| 51 |
+
"should_log": false,
|
| 52 |
+
"should_save": true,
|
| 53 |
+
"should_training_stop": false
|
| 54 |
+
},
|
| 55 |
+
"attributes": {}
|
| 56 |
+
}
|
| 57 |
+
},
|
| 58 |
+
"total_flos": 6199671921083520.0,
|
| 59 |
+
"train_batch_size": 32,
|
| 60 |
+
"trial_name": null,
|
| 61 |
+
"trial_params": null
|
| 62 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a369d04c392a152038f999ecba936911bcde812334c182dac021a19c2d915389
|
| 3 |
+
size 5201
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ACCOUNT_NUMBER",
|
| 15 |
+
"2": "ADDRESS",
|
| 16 |
+
"3": "API_KEY",
|
| 17 |
+
"4": "BANK_ROUTING_NUMBER",
|
| 18 |
+
"5": "BIOMETRIC_IDENTIFIER",
|
| 19 |
+
"6": "CERTIFICATE_LICENSE_NUMBER",
|
| 20 |
+
"7": "CITY",
|
| 21 |
+
"8": "COMPANY_NAME",
|
| 22 |
+
"9": "COORDINATE",
|
| 23 |
+
"10": "COUNTRY",
|
| 24 |
+
"11": "CREDIT_CARD_NUMBER",
|
| 25 |
+
"12": "CUSTOMER_ID",
|
| 26 |
+
"13": "CVV",
|
| 27 |
+
"14": "DATE",
|
| 28 |
+
"15": "DATE_OF_BIRTH",
|
| 29 |
+
"16": "DATE_TIME",
|
| 30 |
+
"17": "DEVICE_IDENTIFIER",
|
| 31 |
+
"18": "EMAIL",
|
| 32 |
+
"19": "EMPLOYEE_ID",
|
| 33 |
+
"20": "FIRST_NAME",
|
| 34 |
+
"21": "HEALTH_PLAN_BENEFICIARY_NUMBER",
|
| 35 |
+
"22": "IPV4",
|
| 36 |
+
"23": "IPV6",
|
| 37 |
+
"24": "LAST_NAME",
|
| 38 |
+
"25": "LICENSE_PLATE",
|
| 39 |
+
"26": "MEDICAL_RECORD_NUMBER",
|
| 40 |
+
"27": "NAME",
|
| 41 |
+
"28": "NATIONAL_ID",
|
| 42 |
+
"29": "PASSWORD",
|
| 43 |
+
"30": "PHONE_NUMBER",
|
| 44 |
+
"31": "PIN",
|
| 45 |
+
"32": "POSTCODE",
|
| 46 |
+
"33": "SSN",
|
| 47 |
+
"34": "STATE",
|
| 48 |
+
"35": "STREET_ADDRESS",
|
| 49 |
+
"36": "SWIFT_BIC",
|
| 50 |
+
"37": "TAX_ID",
|
| 51 |
+
"38": "TIME",
|
| 52 |
+
"39": "UNIQUE_IDENTIFIER",
|
| 53 |
+
"40": "URL",
|
| 54 |
+
"41": "USER_NAME",
|
| 55 |
+
"42": "VEHICLE_IDENTIFIER"
|
| 56 |
+
},
|
| 57 |
+
"initializer_range": 0.02,
|
| 58 |
+
"intermediate_size": 3072,
|
| 59 |
+
"label2id": {
|
| 60 |
+
"ACCOUNT_NUMBER": 1,
|
| 61 |
+
"ADDRESS": 2,
|
| 62 |
+
"API_KEY": 3,
|
| 63 |
+
"BANK_ROUTING_NUMBER": 4,
|
| 64 |
+
"BIOMETRIC_IDENTIFIER": 5,
|
| 65 |
+
"CERTIFICATE_LICENSE_NUMBER": 6,
|
| 66 |
+
"CITY": 7,
|
| 67 |
+
"COMPANY_NAME": 8,
|
| 68 |
+
"COORDINATE": 9,
|
| 69 |
+
"COUNTRY": 10,
|
| 70 |
+
"CREDIT_CARD_NUMBER": 11,
|
| 71 |
+
"CUSTOMER_ID": 12,
|
| 72 |
+
"CVV": 13,
|
| 73 |
+
"DATE": 14,
|
| 74 |
+
"DATE_OF_BIRTH": 15,
|
| 75 |
+
"DATE_TIME": 16,
|
| 76 |
+
"DEVICE_IDENTIFIER": 17,
|
| 77 |
+
"EMAIL": 18,
|
| 78 |
+
"EMPLOYEE_ID": 19,
|
| 79 |
+
"FIRST_NAME": 20,
|
| 80 |
+
"HEALTH_PLAN_BENEFICIARY_NUMBER": 21,
|
| 81 |
+
"IPV4": 22,
|
| 82 |
+
"IPV6": 23,
|
| 83 |
+
"LAST_NAME": 24,
|
| 84 |
+
"LICENSE_PLATE": 25,
|
| 85 |
+
"MEDICAL_RECORD_NUMBER": 26,
|
| 86 |
+
"NAME": 27,
|
| 87 |
+
"NATIONAL_ID": 28,
|
| 88 |
+
"O": 0,
|
| 89 |
+
"PASSWORD": 29,
|
| 90 |
+
"PHONE_NUMBER": 30,
|
| 91 |
+
"PIN": 31,
|
| 92 |
+
"POSTCODE": 32,
|
| 93 |
+
"SSN": 33,
|
| 94 |
+
"STATE": 34,
|
| 95 |
+
"STREET_ADDRESS": 35,
|
| 96 |
+
"SWIFT_BIC": 36,
|
| 97 |
+
"TAX_ID": 37,
|
| 98 |
+
"TIME": 38,
|
| 99 |
+
"UNIQUE_IDENTIFIER": 39,
|
| 100 |
+
"URL": 40,
|
| 101 |
+
"USER_NAME": 41,
|
| 102 |
+
"VEHICLE_IDENTIFIER": 42
|
| 103 |
+
},
|
| 104 |
+
"layer_norm_eps": 1e-07,
|
| 105 |
+
"legacy": true,
|
| 106 |
+
"max_position_embeddings": 512,
|
| 107 |
+
"max_relative_positions": -1,
|
| 108 |
+
"model_type": "deberta-v2",
|
| 109 |
+
"norm_rel_ebd": "layer_norm",
|
| 110 |
+
"num_attention_heads": 12,
|
| 111 |
+
"num_hidden_layers": 12,
|
| 112 |
+
"pad_token_id": 0,
|
| 113 |
+
"pooler_dropout": 0,
|
| 114 |
+
"pooler_hidden_act": "gelu",
|
| 115 |
+
"pooler_hidden_size": 768,
|
| 116 |
+
"pos_att_type": [
|
| 117 |
+
"p2c",
|
| 118 |
+
"c2p"
|
| 119 |
+
],
|
| 120 |
+
"position_biased_input": false,
|
| 121 |
+
"position_buckets": 256,
|
| 122 |
+
"relative_attention": true,
|
| 123 |
+
"share_att_key": true,
|
| 124 |
+
"tie_word_embeddings": true,
|
| 125 |
+
"transformers_version": "5.3.0",
|
| 126 |
+
"type_vocab_size": 0,
|
| 127 |
+
"use_cache": false,
|
| 128 |
+
"vocab_size": 128100
|
| 129 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2711d29ea0ba2edbb85a92a206d7e88a26c0d8d4d716fb5b95737ecb46aefaef
|
| 3 |
+
size 741461252
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5086d34465f5b419b25d27562ce599c78701585cb722ec1300dd2efc41fa1da3
|
| 3 |
+
size 1483046155
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2889a5fc7e1351c254f74295f4963a0087c3299af93fe55710d666aa22c5d9b1
|
| 3 |
+
size 14645
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3e33a32c5abe4f841db7caedbcfcb0bc614df83c1a19d7fd381ca3242065525
|
| 3 |
+
size 1465
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/tokenizer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "[CLS]",
|
| 5 |
+
"cls_token": "[CLS]",
|
| 6 |
+
"do_lower_case": false,
|
| 7 |
+
"eos_token": "[SEP]",
|
| 8 |
+
"extra_special_tokens": [
|
| 9 |
+
"[PAD]",
|
| 10 |
+
"[CLS]",
|
| 11 |
+
"[SEP]"
|
| 12 |
+
],
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"mask_token": "[MASK]",
|
| 15 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 16 |
+
"pad_token": "[PAD]",
|
| 17 |
+
"sep_token": "[SEP]",
|
| 18 |
+
"split_by_punct": false,
|
| 19 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 20 |
+
"unk_id": 3,
|
| 21 |
+
"unk_token": "[UNK]",
|
| 22 |
+
"vocab_type": "spm"
|
| 23 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/trainer_state.json
ADDED
|
@@ -0,0 +1,69 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.8887308922858158,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.17774617845716317,
|
| 14 |
+
"grad_norm": 0.011032640002667904,
|
| 15 |
+
"learning_rate": 1.8226093138997513e-05,
|
| 16 |
+
"loss": 0.007762076377868653,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.35549235691432635,
|
| 21 |
+
"grad_norm": 0.012853773310780525,
|
| 22 |
+
"learning_rate": 1.6448631354425882e-05,
|
| 23 |
+
"loss": 0.0016952511072158814,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5332385353714895,
|
| 28 |
+
"grad_norm": 0.00953384954482317,
|
| 29 |
+
"learning_rate": 1.4671169569854249e-05,
|
| 30 |
+
"loss": 0.0008515777587890625,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.7109847138286527,
|
| 35 |
+
"grad_norm": 0.012823447585105896,
|
| 36 |
+
"learning_rate": 1.2893707785282616e-05,
|
| 37 |
+
"loss": 0.0006071038842201233,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.8887308922858158,
|
| 42 |
+
"grad_norm": 0.008121342398226261,
|
| 43 |
+
"learning_rate": 1.1116246000710985e-05,
|
| 44 |
+
"loss": 0.00048150041699409485,
|
| 45 |
+
"step": 2500
|
| 46 |
+
}
|
| 47 |
+
],
|
| 48 |
+
"logging_steps": 500,
|
| 49 |
+
"max_steps": 5626,
|
| 50 |
+
"num_input_tokens_seen": 0,
|
| 51 |
+
"num_train_epochs": 2,
|
| 52 |
+
"save_steps": 500,
|
| 53 |
+
"stateful_callbacks": {
|
| 54 |
+
"TrainerControl": {
|
| 55 |
+
"args": {
|
| 56 |
+
"should_epoch_stop": false,
|
| 57 |
+
"should_evaluate": false,
|
| 58 |
+
"should_log": false,
|
| 59 |
+
"should_save": true,
|
| 60 |
+
"should_training_stop": false
|
| 61 |
+
},
|
| 62 |
+
"attributes": {}
|
| 63 |
+
}
|
| 64 |
+
},
|
| 65 |
+
"total_flos": 7735287824498304.0,
|
| 66 |
+
"train_batch_size": 32,
|
| 67 |
+
"trial_name": null,
|
| 68 |
+
"trial_params": null
|
| 69 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-2500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a369d04c392a152038f999ecba936911bcde812334c182dac021a19c2d915389
|
| 3 |
+
size 5201
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ACCOUNT_NUMBER",
|
| 15 |
+
"2": "ADDRESS",
|
| 16 |
+
"3": "API_KEY",
|
| 17 |
+
"4": "BANK_ROUTING_NUMBER",
|
| 18 |
+
"5": "BIOMETRIC_IDENTIFIER",
|
| 19 |
+
"6": "CERTIFICATE_LICENSE_NUMBER",
|
| 20 |
+
"7": "CITY",
|
| 21 |
+
"8": "COMPANY_NAME",
|
| 22 |
+
"9": "COORDINATE",
|
| 23 |
+
"10": "COUNTRY",
|
| 24 |
+
"11": "CREDIT_CARD_NUMBER",
|
| 25 |
+
"12": "CUSTOMER_ID",
|
| 26 |
+
"13": "CVV",
|
| 27 |
+
"14": "DATE",
|
| 28 |
+
"15": "DATE_OF_BIRTH",
|
| 29 |
+
"16": "DATE_TIME",
|
| 30 |
+
"17": "DEVICE_IDENTIFIER",
|
| 31 |
+
"18": "EMAIL",
|
| 32 |
+
"19": "EMPLOYEE_ID",
|
| 33 |
+
"20": "FIRST_NAME",
|
| 34 |
+
"21": "HEALTH_PLAN_BENEFICIARY_NUMBER",
|
| 35 |
+
"22": "IPV4",
|
| 36 |
+
"23": "IPV6",
|
| 37 |
+
"24": "LAST_NAME",
|
| 38 |
+
"25": "LICENSE_PLATE",
|
| 39 |
+
"26": "MEDICAL_RECORD_NUMBER",
|
| 40 |
+
"27": "NAME",
|
| 41 |
+
"28": "NATIONAL_ID",
|
| 42 |
+
"29": "PASSWORD",
|
| 43 |
+
"30": "PHONE_NUMBER",
|
| 44 |
+
"31": "PIN",
|
| 45 |
+
"32": "POSTCODE",
|
| 46 |
+
"33": "SSN",
|
| 47 |
+
"34": "STATE",
|
| 48 |
+
"35": "STREET_ADDRESS",
|
| 49 |
+
"36": "SWIFT_BIC",
|
| 50 |
+
"37": "TAX_ID",
|
| 51 |
+
"38": "TIME",
|
| 52 |
+
"39": "UNIQUE_IDENTIFIER",
|
| 53 |
+
"40": "URL",
|
| 54 |
+
"41": "USER_NAME",
|
| 55 |
+
"42": "VEHICLE_IDENTIFIER"
|
| 56 |
+
},
|
| 57 |
+
"initializer_range": 0.02,
|
| 58 |
+
"intermediate_size": 3072,
|
| 59 |
+
"label2id": {
|
| 60 |
+
"ACCOUNT_NUMBER": 1,
|
| 61 |
+
"ADDRESS": 2,
|
| 62 |
+
"API_KEY": 3,
|
| 63 |
+
"BANK_ROUTING_NUMBER": 4,
|
| 64 |
+
"BIOMETRIC_IDENTIFIER": 5,
|
| 65 |
+
"CERTIFICATE_LICENSE_NUMBER": 6,
|
| 66 |
+
"CITY": 7,
|
| 67 |
+
"COMPANY_NAME": 8,
|
| 68 |
+
"COORDINATE": 9,
|
| 69 |
+
"COUNTRY": 10,
|
| 70 |
+
"CREDIT_CARD_NUMBER": 11,
|
| 71 |
+
"CUSTOMER_ID": 12,
|
| 72 |
+
"CVV": 13,
|
| 73 |
+
"DATE": 14,
|
| 74 |
+
"DATE_OF_BIRTH": 15,
|
| 75 |
+
"DATE_TIME": 16,
|
| 76 |
+
"DEVICE_IDENTIFIER": 17,
|
| 77 |
+
"EMAIL": 18,
|
| 78 |
+
"EMPLOYEE_ID": 19,
|
| 79 |
+
"FIRST_NAME": 20,
|
| 80 |
+
"HEALTH_PLAN_BENEFICIARY_NUMBER": 21,
|
| 81 |
+
"IPV4": 22,
|
| 82 |
+
"IPV6": 23,
|
| 83 |
+
"LAST_NAME": 24,
|
| 84 |
+
"LICENSE_PLATE": 25,
|
| 85 |
+
"MEDICAL_RECORD_NUMBER": 26,
|
| 86 |
+
"NAME": 27,
|
| 87 |
+
"NATIONAL_ID": 28,
|
| 88 |
+
"O": 0,
|
| 89 |
+
"PASSWORD": 29,
|
| 90 |
+
"PHONE_NUMBER": 30,
|
| 91 |
+
"PIN": 31,
|
| 92 |
+
"POSTCODE": 32,
|
| 93 |
+
"SSN": 33,
|
| 94 |
+
"STATE": 34,
|
| 95 |
+
"STREET_ADDRESS": 35,
|
| 96 |
+
"SWIFT_BIC": 36,
|
| 97 |
+
"TAX_ID": 37,
|
| 98 |
+
"TIME": 38,
|
| 99 |
+
"UNIQUE_IDENTIFIER": 39,
|
| 100 |
+
"URL": 40,
|
| 101 |
+
"USER_NAME": 41,
|
| 102 |
+
"VEHICLE_IDENTIFIER": 42
|
| 103 |
+
},
|
| 104 |
+
"layer_norm_eps": 1e-07,
|
| 105 |
+
"legacy": true,
|
| 106 |
+
"max_position_embeddings": 512,
|
| 107 |
+
"max_relative_positions": -1,
|
| 108 |
+
"model_type": "deberta-v2",
|
| 109 |
+
"norm_rel_ebd": "layer_norm",
|
| 110 |
+
"num_attention_heads": 12,
|
| 111 |
+
"num_hidden_layers": 12,
|
| 112 |
+
"pad_token_id": 0,
|
| 113 |
+
"pooler_dropout": 0,
|
| 114 |
+
"pooler_hidden_act": "gelu",
|
| 115 |
+
"pooler_hidden_size": 768,
|
| 116 |
+
"pos_att_type": [
|
| 117 |
+
"p2c",
|
| 118 |
+
"c2p"
|
| 119 |
+
],
|
| 120 |
+
"position_biased_input": false,
|
| 121 |
+
"position_buckets": 256,
|
| 122 |
+
"relative_attention": true,
|
| 123 |
+
"share_att_key": true,
|
| 124 |
+
"tie_word_embeddings": true,
|
| 125 |
+
"transformers_version": "5.3.0",
|
| 126 |
+
"type_vocab_size": 0,
|
| 127 |
+
"use_cache": false,
|
| 128 |
+
"vocab_size": 128100
|
| 129 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02afa038148645ddcf500c4cea6fd8e2f71c789644a926543ec75bd4c2bb6fb4
|
| 3 |
+
size 741461252
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b18b8f082ac9ecd33607869ee857408f407aff8406b8e0d3cb5225e906a05b4f
|
| 3 |
+
size 1483046155
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:26cafe65115d84ed23c671ddbb18f1c706764408dcdc640d291d6d0f3fe9567f
|
| 3 |
+
size 14645
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dc8794b7276022912e513cb31c43010f7261560910d8a0eb75679a992b60210d
|
| 3 |
+
size 1465
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "[CLS]",
|
| 5 |
+
"cls_token": "[CLS]",
|
| 6 |
+
"do_lower_case": false,
|
| 7 |
+
"eos_token": "[SEP]",
|
| 8 |
+
"extra_special_tokens": [
|
| 9 |
+
"[PAD]",
|
| 10 |
+
"[CLS]",
|
| 11 |
+
"[SEP]"
|
| 12 |
+
],
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"mask_token": "[MASK]",
|
| 15 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 16 |
+
"pad_token": "[PAD]",
|
| 17 |
+
"sep_token": "[SEP]",
|
| 18 |
+
"split_by_punct": false,
|
| 19 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 20 |
+
"unk_id": 3,
|
| 21 |
+
"unk_token": "[UNK]",
|
| 22 |
+
"vocab_type": "spm"
|
| 23 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/trainer_state.json
ADDED
|
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.066477070742979,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.17774617845716317,
|
| 14 |
+
"grad_norm": 0.011032640002667904,
|
| 15 |
+
"learning_rate": 1.8226093138997513e-05,
|
| 16 |
+
"loss": 0.007762076377868653,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.35549235691432635,
|
| 21 |
+
"grad_norm": 0.012853773310780525,
|
| 22 |
+
"learning_rate": 1.6448631354425882e-05,
|
| 23 |
+
"loss": 0.0016952511072158814,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5332385353714895,
|
| 28 |
+
"grad_norm": 0.00953384954482317,
|
| 29 |
+
"learning_rate": 1.4671169569854249e-05,
|
| 30 |
+
"loss": 0.0008515777587890625,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.7109847138286527,
|
| 35 |
+
"grad_norm": 0.012823447585105896,
|
| 36 |
+
"learning_rate": 1.2893707785282616e-05,
|
| 37 |
+
"loss": 0.0006071038842201233,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.8887308922858158,
|
| 42 |
+
"grad_norm": 0.008121342398226261,
|
| 43 |
+
"learning_rate": 1.1116246000710985e-05,
|
| 44 |
+
"loss": 0.00048150041699409485,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.066477070742979,
|
| 49 |
+
"grad_norm": 0.005901801865547895,
|
| 50 |
+
"learning_rate": 9.338784216139354e-06,
|
| 51 |
+
"loss": 0.00044361612200737,
|
| 52 |
+
"step": 3000
|
| 53 |
+
}
|
| 54 |
+
],
|
| 55 |
+
"logging_steps": 500,
|
| 56 |
+
"max_steps": 5626,
|
| 57 |
+
"num_input_tokens_seen": 0,
|
| 58 |
+
"num_train_epochs": 2,
|
| 59 |
+
"save_steps": 500,
|
| 60 |
+
"stateful_callbacks": {
|
| 61 |
+
"TrainerControl": {
|
| 62 |
+
"args": {
|
| 63 |
+
"should_epoch_stop": false,
|
| 64 |
+
"should_evaluate": false,
|
| 65 |
+
"should_log": false,
|
| 66 |
+
"should_save": true,
|
| 67 |
+
"should_training_stop": false
|
| 68 |
+
},
|
| 69 |
+
"attributes": {}
|
| 70 |
+
}
|
| 71 |
+
},
|
| 72 |
+
"total_flos": 9273853894841568.0,
|
| 73 |
+
"train_batch_size": 32,
|
| 74 |
+
"trial_name": null,
|
| 75 |
+
"trial_params": null
|
| 76 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a369d04c392a152038f999ecba936911bcde812334c182dac021a19c2d915389
|
| 3 |
+
size 5201
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ACCOUNT_NUMBER",
|
| 15 |
+
"2": "ADDRESS",
|
| 16 |
+
"3": "API_KEY",
|
| 17 |
+
"4": "BANK_ROUTING_NUMBER",
|
| 18 |
+
"5": "BIOMETRIC_IDENTIFIER",
|
| 19 |
+
"6": "CERTIFICATE_LICENSE_NUMBER",
|
| 20 |
+
"7": "CITY",
|
| 21 |
+
"8": "COMPANY_NAME",
|
| 22 |
+
"9": "COORDINATE",
|
| 23 |
+
"10": "COUNTRY",
|
| 24 |
+
"11": "CREDIT_CARD_NUMBER",
|
| 25 |
+
"12": "CUSTOMER_ID",
|
| 26 |
+
"13": "CVV",
|
| 27 |
+
"14": "DATE",
|
| 28 |
+
"15": "DATE_OF_BIRTH",
|
| 29 |
+
"16": "DATE_TIME",
|
| 30 |
+
"17": "DEVICE_IDENTIFIER",
|
| 31 |
+
"18": "EMAIL",
|
| 32 |
+
"19": "EMPLOYEE_ID",
|
| 33 |
+
"20": "FIRST_NAME",
|
| 34 |
+
"21": "HEALTH_PLAN_BENEFICIARY_NUMBER",
|
| 35 |
+
"22": "IPV4",
|
| 36 |
+
"23": "IPV6",
|
| 37 |
+
"24": "LAST_NAME",
|
| 38 |
+
"25": "LICENSE_PLATE",
|
| 39 |
+
"26": "MEDICAL_RECORD_NUMBER",
|
| 40 |
+
"27": "NAME",
|
| 41 |
+
"28": "NATIONAL_ID",
|
| 42 |
+
"29": "PASSWORD",
|
| 43 |
+
"30": "PHONE_NUMBER",
|
| 44 |
+
"31": "PIN",
|
| 45 |
+
"32": "POSTCODE",
|
| 46 |
+
"33": "SSN",
|
| 47 |
+
"34": "STATE",
|
| 48 |
+
"35": "STREET_ADDRESS",
|
| 49 |
+
"36": "SWIFT_BIC",
|
| 50 |
+
"37": "TAX_ID",
|
| 51 |
+
"38": "TIME",
|
| 52 |
+
"39": "UNIQUE_IDENTIFIER",
|
| 53 |
+
"40": "URL",
|
| 54 |
+
"41": "USER_NAME",
|
| 55 |
+
"42": "VEHICLE_IDENTIFIER"
|
| 56 |
+
},
|
| 57 |
+
"initializer_range": 0.02,
|
| 58 |
+
"intermediate_size": 3072,
|
| 59 |
+
"label2id": {
|
| 60 |
+
"ACCOUNT_NUMBER": 1,
|
| 61 |
+
"ADDRESS": 2,
|
| 62 |
+
"API_KEY": 3,
|
| 63 |
+
"BANK_ROUTING_NUMBER": 4,
|
| 64 |
+
"BIOMETRIC_IDENTIFIER": 5,
|
| 65 |
+
"CERTIFICATE_LICENSE_NUMBER": 6,
|
| 66 |
+
"CITY": 7,
|
| 67 |
+
"COMPANY_NAME": 8,
|
| 68 |
+
"COORDINATE": 9,
|
| 69 |
+
"COUNTRY": 10,
|
| 70 |
+
"CREDIT_CARD_NUMBER": 11,
|
| 71 |
+
"CUSTOMER_ID": 12,
|
| 72 |
+
"CVV": 13,
|
| 73 |
+
"DATE": 14,
|
| 74 |
+
"DATE_OF_BIRTH": 15,
|
| 75 |
+
"DATE_TIME": 16,
|
| 76 |
+
"DEVICE_IDENTIFIER": 17,
|
| 77 |
+
"EMAIL": 18,
|
| 78 |
+
"EMPLOYEE_ID": 19,
|
| 79 |
+
"FIRST_NAME": 20,
|
| 80 |
+
"HEALTH_PLAN_BENEFICIARY_NUMBER": 21,
|
| 81 |
+
"IPV4": 22,
|
| 82 |
+
"IPV6": 23,
|
| 83 |
+
"LAST_NAME": 24,
|
| 84 |
+
"LICENSE_PLATE": 25,
|
| 85 |
+
"MEDICAL_RECORD_NUMBER": 26,
|
| 86 |
+
"NAME": 27,
|
| 87 |
+
"NATIONAL_ID": 28,
|
| 88 |
+
"O": 0,
|
| 89 |
+
"PASSWORD": 29,
|
| 90 |
+
"PHONE_NUMBER": 30,
|
| 91 |
+
"PIN": 31,
|
| 92 |
+
"POSTCODE": 32,
|
| 93 |
+
"SSN": 33,
|
| 94 |
+
"STATE": 34,
|
| 95 |
+
"STREET_ADDRESS": 35,
|
| 96 |
+
"SWIFT_BIC": 36,
|
| 97 |
+
"TAX_ID": 37,
|
| 98 |
+
"TIME": 38,
|
| 99 |
+
"UNIQUE_IDENTIFIER": 39,
|
| 100 |
+
"URL": 40,
|
| 101 |
+
"USER_NAME": 41,
|
| 102 |
+
"VEHICLE_IDENTIFIER": 42
|
| 103 |
+
},
|
| 104 |
+
"layer_norm_eps": 1e-07,
|
| 105 |
+
"legacy": true,
|
| 106 |
+
"max_position_embeddings": 512,
|
| 107 |
+
"max_relative_positions": -1,
|
| 108 |
+
"model_type": "deberta-v2",
|
| 109 |
+
"norm_rel_ebd": "layer_norm",
|
| 110 |
+
"num_attention_heads": 12,
|
| 111 |
+
"num_hidden_layers": 12,
|
| 112 |
+
"pad_token_id": 0,
|
| 113 |
+
"pooler_dropout": 0,
|
| 114 |
+
"pooler_hidden_act": "gelu",
|
| 115 |
+
"pooler_hidden_size": 768,
|
| 116 |
+
"pos_att_type": [
|
| 117 |
+
"p2c",
|
| 118 |
+
"c2p"
|
| 119 |
+
],
|
| 120 |
+
"position_biased_input": false,
|
| 121 |
+
"position_buckets": 256,
|
| 122 |
+
"relative_attention": true,
|
| 123 |
+
"share_att_key": true,
|
| 124 |
+
"tie_word_embeddings": true,
|
| 125 |
+
"transformers_version": "5.3.0",
|
| 126 |
+
"type_vocab_size": 0,
|
| 127 |
+
"use_cache": false,
|
| 128 |
+
"vocab_size": 128100
|
| 129 |
+
}
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:caba2dcb62a45f90ea6b0d1e63ce37bcebbf4b1530f8c0fac0cd88b15d57207c
|
| 3 |
+
size 741461252
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:54977ac9a7c4ed7b42de1b0e81e81e97e76773c32c5643d050915a121ceacfc7
|
| 3 |
+
size 1483046155
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee6381d9d4115c0977677057a84fe89e8fc1c4a61ba0c4eddb59dfe6f25dba22
|
| 3 |
+
size 14645
|
gretel-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-3500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c91b301817590bbdb8afcc102ef3f6a29cebdbe88f8ba0e843de0bb61fc01cec
|
| 3 |
+
size 1465
|