Upload folder using huggingface_hub
Browse files- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/config.json +157 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/model.safetensors +3 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/optimizer.pt +3 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/rng_state.pth +3 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/scaler.pt +3 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/scheduler.pt +3 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/tokenizer.json +0 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/tokenizer_config.json +23 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/trainer_state.json +181 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/training_args.bin +3 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/config.json +157 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/pytorch_model.bin +3 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/tokenizer.json +0 -0
- pii-masking-200k/SpanBased-CustomDeBERTa/tokenizer_config.json +23 -0
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/config.json
ADDED
|
@@ -0,0 +1,157 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "ACCOUNTNAME",
|
| 15 |
+
"2": "ACCOUNTNUMBER",
|
| 16 |
+
"3": "AGE",
|
| 17 |
+
"4": "AMOUNT",
|
| 18 |
+
"5": "BIC",
|
| 19 |
+
"6": "BITCOINADDRESS",
|
| 20 |
+
"7": "BUILDINGNUMBER",
|
| 21 |
+
"8": "CITY",
|
| 22 |
+
"9": "COMPANYNAME",
|
| 23 |
+
"10": "COUNTY",
|
| 24 |
+
"11": "CREDITCARDCVV",
|
| 25 |
+
"12": "CREDITCARDISSUER",
|
| 26 |
+
"13": "CREDITCARDNUMBER",
|
| 27 |
+
"14": "CURRENCY",
|
| 28 |
+
"15": "CURRENCYCODE",
|
| 29 |
+
"16": "CURRENCYNAME",
|
| 30 |
+
"17": "CURRENCYSYMBOL",
|
| 31 |
+
"18": "DATE",
|
| 32 |
+
"19": "DOB",
|
| 33 |
+
"20": "EMAIL",
|
| 34 |
+
"21": "ETHEREUMADDRESS",
|
| 35 |
+
"22": "EYECOLOR",
|
| 36 |
+
"23": "FIRSTNAME",
|
| 37 |
+
"24": "GENDER",
|
| 38 |
+
"25": "HEIGHT",
|
| 39 |
+
"26": "IBAN",
|
| 40 |
+
"27": "IP",
|
| 41 |
+
"28": "IPV4",
|
| 42 |
+
"29": "IPV6",
|
| 43 |
+
"30": "JOBAREA",
|
| 44 |
+
"31": "JOBTITLE",
|
| 45 |
+
"32": "JOBTYPE",
|
| 46 |
+
"33": "LASTNAME",
|
| 47 |
+
"34": "LITECOINADDRESS",
|
| 48 |
+
"35": "MAC",
|
| 49 |
+
"36": "MASKEDNUMBER",
|
| 50 |
+
"37": "MIDDLENAME",
|
| 51 |
+
"38": "NEARBYGPSCOORDINATE",
|
| 52 |
+
"39": "ORDINALDIRECTION",
|
| 53 |
+
"40": "PASSWORD",
|
| 54 |
+
"41": "PHONEIMEI",
|
| 55 |
+
"42": "PHONENUMBER",
|
| 56 |
+
"43": "PIN",
|
| 57 |
+
"44": "PREFIX",
|
| 58 |
+
"45": "SECONDARYADDRESS",
|
| 59 |
+
"46": "SEX",
|
| 60 |
+
"47": "SSN",
|
| 61 |
+
"48": "STATE",
|
| 62 |
+
"49": "STREET",
|
| 63 |
+
"50": "TIME",
|
| 64 |
+
"51": "URL",
|
| 65 |
+
"52": "USERAGENT",
|
| 66 |
+
"53": "USERNAME",
|
| 67 |
+
"54": "VEHICLEVIN",
|
| 68 |
+
"55": "VEHICLEVRM",
|
| 69 |
+
"56": "ZIPCODE"
|
| 70 |
+
},
|
| 71 |
+
"initializer_range": 0.02,
|
| 72 |
+
"intermediate_size": 3072,
|
| 73 |
+
"label2id": {
|
| 74 |
+
"ACCOUNTNAME": 1,
|
| 75 |
+
"ACCOUNTNUMBER": 2,
|
| 76 |
+
"AGE": 3,
|
| 77 |
+
"AMOUNT": 4,
|
| 78 |
+
"BIC": 5,
|
| 79 |
+
"BITCOINADDRESS": 6,
|
| 80 |
+
"BUILDINGNUMBER": 7,
|
| 81 |
+
"CITY": 8,
|
| 82 |
+
"COMPANYNAME": 9,
|
| 83 |
+
"COUNTY": 10,
|
| 84 |
+
"CREDITCARDCVV": 11,
|
| 85 |
+
"CREDITCARDISSUER": 12,
|
| 86 |
+
"CREDITCARDNUMBER": 13,
|
| 87 |
+
"CURRENCY": 14,
|
| 88 |
+
"CURRENCYCODE": 15,
|
| 89 |
+
"CURRENCYNAME": 16,
|
| 90 |
+
"CURRENCYSYMBOL": 17,
|
| 91 |
+
"DATE": 18,
|
| 92 |
+
"DOB": 19,
|
| 93 |
+
"EMAIL": 20,
|
| 94 |
+
"ETHEREUMADDRESS": 21,
|
| 95 |
+
"EYECOLOR": 22,
|
| 96 |
+
"FIRSTNAME": 23,
|
| 97 |
+
"GENDER": 24,
|
| 98 |
+
"HEIGHT": 25,
|
| 99 |
+
"IBAN": 26,
|
| 100 |
+
"IP": 27,
|
| 101 |
+
"IPV4": 28,
|
| 102 |
+
"IPV6": 29,
|
| 103 |
+
"JOBAREA": 30,
|
| 104 |
+
"JOBTITLE": 31,
|
| 105 |
+
"JOBTYPE": 32,
|
| 106 |
+
"LASTNAME": 33,
|
| 107 |
+
"LITECOINADDRESS": 34,
|
| 108 |
+
"MAC": 35,
|
| 109 |
+
"MASKEDNUMBER": 36,
|
| 110 |
+
"MIDDLENAME": 37,
|
| 111 |
+
"NEARBYGPSCOORDINATE": 38,
|
| 112 |
+
"O": 0,
|
| 113 |
+
"ORDINALDIRECTION": 39,
|
| 114 |
+
"PASSWORD": 40,
|
| 115 |
+
"PHONEIMEI": 41,
|
| 116 |
+
"PHONENUMBER": 42,
|
| 117 |
+
"PIN": 43,
|
| 118 |
+
"PREFIX": 44,
|
| 119 |
+
"SECONDARYADDRESS": 45,
|
| 120 |
+
"SEX": 46,
|
| 121 |
+
"SSN": 47,
|
| 122 |
+
"STATE": 48,
|
| 123 |
+
"STREET": 49,
|
| 124 |
+
"TIME": 50,
|
| 125 |
+
"URL": 51,
|
| 126 |
+
"USERAGENT": 52,
|
| 127 |
+
"USERNAME": 53,
|
| 128 |
+
"VEHICLEVIN": 54,
|
| 129 |
+
"VEHICLEVRM": 55,
|
| 130 |
+
"ZIPCODE": 56
|
| 131 |
+
},
|
| 132 |
+
"layer_norm_eps": 1e-07,
|
| 133 |
+
"legacy": true,
|
| 134 |
+
"max_position_embeddings": 512,
|
| 135 |
+
"max_relative_positions": -1,
|
| 136 |
+
"model_type": "deberta-v2",
|
| 137 |
+
"norm_rel_ebd": "layer_norm",
|
| 138 |
+
"num_attention_heads": 12,
|
| 139 |
+
"num_hidden_layers": 12,
|
| 140 |
+
"pad_token_id": 0,
|
| 141 |
+
"pooler_dropout": 0,
|
| 142 |
+
"pooler_hidden_act": "gelu",
|
| 143 |
+
"pooler_hidden_size": 768,
|
| 144 |
+
"pos_att_type": [
|
| 145 |
+
"p2c",
|
| 146 |
+
"c2p"
|
| 147 |
+
],
|
| 148 |
+
"position_biased_input": false,
|
| 149 |
+
"position_buckets": 256,
|
| 150 |
+
"relative_attention": true,
|
| 151 |
+
"share_att_key": true,
|
| 152 |
+
"tie_word_embeddings": true,
|
| 153 |
+
"transformers_version": "5.3.0",
|
| 154 |
+
"type_vocab_size": 0,
|
| 155 |
+
"use_cache": false,
|
| 156 |
+
"vocab_size": 128100
|
| 157 |
+
}
|
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5813aee93a6c9c72d0a578ffbcb62c14a39733f40b691825b96eddc10298072e
|
| 3 |
+
size 741504316
|
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cb913daeb5b02650eb8ce0d2f55bd336c331cb68e9e5b015ffdd285ba01ba404
|
| 3 |
+
size 1483132299
|
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f32b1cdf76acb83e0ab16eb992b305f0052721f6c55cce1dc16704430c086ac
|
| 3 |
+
size 14645
|
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:148ca1e14023632ea13349bbcd8f280021f5bd8fa861b06e649de22dcb890e71
|
| 3 |
+
size 1383
|
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50fd2a9464797225dd77fdcda28cb6d4a004edb05ae6a957c4f50df32d15cc8b
|
| 3 |
+
size 1465
|
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/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 |
+
}
|
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/trainer_state.json
ADDED
|
@@ -0,0 +1,181 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 10838,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.09226794611551947,
|
| 14 |
+
"grad_norm": 0.013377240858972073,
|
| 15 |
+
"learning_rate": 1.9079165897767117e-05,
|
| 16 |
+
"loss": 0.011818027496337891,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.18453589223103894,
|
| 21 |
+
"grad_norm": 0.01368043664842844,
|
| 22 |
+
"learning_rate": 1.815648643661192e-05,
|
| 23 |
+
"loss": 0.0040970339775085445,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2768038383465584,
|
| 28 |
+
"grad_norm": 0.015748485922813416,
|
| 29 |
+
"learning_rate": 1.723380697545673e-05,
|
| 30 |
+
"loss": 0.0019339693784713745,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.3690717844620779,
|
| 35 |
+
"grad_norm": 0.013551352545619011,
|
| 36 |
+
"learning_rate": 1.6311127514301534e-05,
|
| 37 |
+
"loss": 0.001239347219467163,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.46133973057759736,
|
| 42 |
+
"grad_norm": 0.015444819815456867,
|
| 43 |
+
"learning_rate": 1.5388448053146338e-05,
|
| 44 |
+
"loss": 0.0009453317523002624,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.5536076766931168,
|
| 49 |
+
"grad_norm": 0.012576445005834103,
|
| 50 |
+
"learning_rate": 1.4465768591991143e-05,
|
| 51 |
+
"loss": 0.0007971944808959961,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.6458756228086363,
|
| 56 |
+
"grad_norm": 0.009161020629107952,
|
| 57 |
+
"learning_rate": 1.3543089130835949e-05,
|
| 58 |
+
"loss": 0.0006877148151397705,
|
| 59 |
+
"step": 3500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.7381435689241558,
|
| 63 |
+
"grad_norm": 0.021290592849254608,
|
| 64 |
+
"learning_rate": 1.2620409669680753e-05,
|
| 65 |
+
"loss": 0.0006193248629570008,
|
| 66 |
+
"step": 4000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.8304115150396753,
|
| 70 |
+
"grad_norm": 0.007458180654793978,
|
| 71 |
+
"learning_rate": 1.169773020852556e-05,
|
| 72 |
+
"loss": 0.0005676882863044738,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.9226794611551947,
|
| 77 |
+
"grad_norm": 0.012828211300075054,
|
| 78 |
+
"learning_rate": 1.0775050747370364e-05,
|
| 79 |
+
"loss": 0.0005238401889801025,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.014947407270714,
|
| 84 |
+
"grad_norm": 0.016309715807437897,
|
| 85 |
+
"learning_rate": 9.85237128621517e-06,
|
| 86 |
+
"loss": 0.0004752824902534485,
|
| 87 |
+
"step": 5500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.1072153533862337,
|
| 91 |
+
"grad_norm": 0.016130661591887474,
|
| 92 |
+
"learning_rate": 8.929691825059975e-06,
|
| 93 |
+
"loss": 0.0004253955185413361,
|
| 94 |
+
"step": 6000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.199483299501753,
|
| 98 |
+
"grad_norm": 0.00841038953512907,
|
| 99 |
+
"learning_rate": 8.007012363904781e-06,
|
| 100 |
+
"loss": 0.00040376508235931397,
|
| 101 |
+
"step": 6500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.2917512456172726,
|
| 105 |
+
"grad_norm": 0.012786151841282845,
|
| 106 |
+
"learning_rate": 7.084332902749585e-06,
|
| 107 |
+
"loss": 0.00037746840715408323,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.384019191732792,
|
| 112 |
+
"grad_norm": 0.020556332543492317,
|
| 113 |
+
"learning_rate": 6.161653441594391e-06,
|
| 114 |
+
"loss": 0.00036286091804504396,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.4762871378483116,
|
| 119 |
+
"grad_norm": 0.002548969816416502,
|
| 120 |
+
"learning_rate": 5.238973980439196e-06,
|
| 121 |
+
"loss": 0.00034481051564216615,
|
| 122 |
+
"step": 8000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.568555083963831,
|
| 126 |
+
"grad_norm": 0.004770679399371147,
|
| 127 |
+
"learning_rate": 4.3162945192840015e-06,
|
| 128 |
+
"loss": 0.0003391757905483246,
|
| 129 |
+
"step": 8500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.6608230300793503,
|
| 133 |
+
"grad_norm": 0.016097497195005417,
|
| 134 |
+
"learning_rate": 3.3936150581288064e-06,
|
| 135 |
+
"loss": 0.0003458640575408936,
|
| 136 |
+
"step": 9000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.7530909761948699,
|
| 140 |
+
"grad_norm": 0.01792999543249607,
|
| 141 |
+
"learning_rate": 2.4709355969736113e-06,
|
| 142 |
+
"loss": 0.00032689249515533445,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.8453589223103894,
|
| 147 |
+
"grad_norm": 0.009682539850473404,
|
| 148 |
+
"learning_rate": 1.5482561358184167e-06,
|
| 149 |
+
"loss": 0.00031686127185821534,
|
| 150 |
+
"step": 10000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.9376268684259088,
|
| 154 |
+
"grad_norm": 0.0033965129405260086,
|
| 155 |
+
"learning_rate": 6.25576674663222e-07,
|
| 156 |
+
"loss": 0.00031349530816078184,
|
| 157 |
+
"step": 10500
|
| 158 |
+
}
|
| 159 |
+
],
|
| 160 |
+
"logging_steps": 500,
|
| 161 |
+
"max_steps": 10838,
|
| 162 |
+
"num_input_tokens_seen": 0,
|
| 163 |
+
"num_train_epochs": 2,
|
| 164 |
+
"save_steps": 500,
|
| 165 |
+
"stateful_callbacks": {
|
| 166 |
+
"TrainerControl": {
|
| 167 |
+
"args": {
|
| 168 |
+
"should_epoch_stop": false,
|
| 169 |
+
"should_evaluate": false,
|
| 170 |
+
"should_log": false,
|
| 171 |
+
"should_save": true,
|
| 172 |
+
"should_training_stop": true
|
| 173 |
+
},
|
| 174 |
+
"attributes": {}
|
| 175 |
+
}
|
| 176 |
+
},
|
| 177 |
+
"total_flos": 8974533831191028.0,
|
| 178 |
+
"train_batch_size": 16,
|
| 179 |
+
"trial_name": null,
|
| 180 |
+
"trial_params": null
|
| 181 |
+
}
|
pii-masking-200k/SpanBased-CustomDeBERTa/checkpoint-10838/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:472edcd837d3b5ec587300c8326149d94c53b9568d213ec51b6dbfa20dce3f9c
|
| 3 |
+
size 5201
|
pii-masking-200k/SpanBased-CustomDeBERTa/config.json
ADDED
|
@@ -0,0 +1,157 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "ACCOUNTNAME",
|
| 15 |
+
"2": "ACCOUNTNUMBER",
|
| 16 |
+
"3": "AGE",
|
| 17 |
+
"4": "AMOUNT",
|
| 18 |
+
"5": "BIC",
|
| 19 |
+
"6": "BITCOINADDRESS",
|
| 20 |
+
"7": "BUILDINGNUMBER",
|
| 21 |
+
"8": "CITY",
|
| 22 |
+
"9": "COMPANYNAME",
|
| 23 |
+
"10": "COUNTY",
|
| 24 |
+
"11": "CREDITCARDCVV",
|
| 25 |
+
"12": "CREDITCARDISSUER",
|
| 26 |
+
"13": "CREDITCARDNUMBER",
|
| 27 |
+
"14": "CURRENCY",
|
| 28 |
+
"15": "CURRENCYCODE",
|
| 29 |
+
"16": "CURRENCYNAME",
|
| 30 |
+
"17": "CURRENCYSYMBOL",
|
| 31 |
+
"18": "DATE",
|
| 32 |
+
"19": "DOB",
|
| 33 |
+
"20": "EMAIL",
|
| 34 |
+
"21": "ETHEREUMADDRESS",
|
| 35 |
+
"22": "EYECOLOR",
|
| 36 |
+
"23": "FIRSTNAME",
|
| 37 |
+
"24": "GENDER",
|
| 38 |
+
"25": "HEIGHT",
|
| 39 |
+
"26": "IBAN",
|
| 40 |
+
"27": "IP",
|
| 41 |
+
"28": "IPV4",
|
| 42 |
+
"29": "IPV6",
|
| 43 |
+
"30": "JOBAREA",
|
| 44 |
+
"31": "JOBTITLE",
|
| 45 |
+
"32": "JOBTYPE",
|
| 46 |
+
"33": "LASTNAME",
|
| 47 |
+
"34": "LITECOINADDRESS",
|
| 48 |
+
"35": "MAC",
|
| 49 |
+
"36": "MASKEDNUMBER",
|
| 50 |
+
"37": "MIDDLENAME",
|
| 51 |
+
"38": "NEARBYGPSCOORDINATE",
|
| 52 |
+
"39": "ORDINALDIRECTION",
|
| 53 |
+
"40": "PASSWORD",
|
| 54 |
+
"41": "PHONEIMEI",
|
| 55 |
+
"42": "PHONENUMBER",
|
| 56 |
+
"43": "PIN",
|
| 57 |
+
"44": "PREFIX",
|
| 58 |
+
"45": "SECONDARYADDRESS",
|
| 59 |
+
"46": "SEX",
|
| 60 |
+
"47": "SSN",
|
| 61 |
+
"48": "STATE",
|
| 62 |
+
"49": "STREET",
|
| 63 |
+
"50": "TIME",
|
| 64 |
+
"51": "URL",
|
| 65 |
+
"52": "USERAGENT",
|
| 66 |
+
"53": "USERNAME",
|
| 67 |
+
"54": "VEHICLEVIN",
|
| 68 |
+
"55": "VEHICLEVRM",
|
| 69 |
+
"56": "ZIPCODE"
|
| 70 |
+
},
|
| 71 |
+
"initializer_range": 0.02,
|
| 72 |
+
"intermediate_size": 3072,
|
| 73 |
+
"label2id": {
|
| 74 |
+
"ACCOUNTNAME": 1,
|
| 75 |
+
"ACCOUNTNUMBER": 2,
|
| 76 |
+
"AGE": 3,
|
| 77 |
+
"AMOUNT": 4,
|
| 78 |
+
"BIC": 5,
|
| 79 |
+
"BITCOINADDRESS": 6,
|
| 80 |
+
"BUILDINGNUMBER": 7,
|
| 81 |
+
"CITY": 8,
|
| 82 |
+
"COMPANYNAME": 9,
|
| 83 |
+
"COUNTY": 10,
|
| 84 |
+
"CREDITCARDCVV": 11,
|
| 85 |
+
"CREDITCARDISSUER": 12,
|
| 86 |
+
"CREDITCARDNUMBER": 13,
|
| 87 |
+
"CURRENCY": 14,
|
| 88 |
+
"CURRENCYCODE": 15,
|
| 89 |
+
"CURRENCYNAME": 16,
|
| 90 |
+
"CURRENCYSYMBOL": 17,
|
| 91 |
+
"DATE": 18,
|
| 92 |
+
"DOB": 19,
|
| 93 |
+
"EMAIL": 20,
|
| 94 |
+
"ETHEREUMADDRESS": 21,
|
| 95 |
+
"EYECOLOR": 22,
|
| 96 |
+
"FIRSTNAME": 23,
|
| 97 |
+
"GENDER": 24,
|
| 98 |
+
"HEIGHT": 25,
|
| 99 |
+
"IBAN": 26,
|
| 100 |
+
"IP": 27,
|
| 101 |
+
"IPV4": 28,
|
| 102 |
+
"IPV6": 29,
|
| 103 |
+
"JOBAREA": 30,
|
| 104 |
+
"JOBTITLE": 31,
|
| 105 |
+
"JOBTYPE": 32,
|
| 106 |
+
"LASTNAME": 33,
|
| 107 |
+
"LITECOINADDRESS": 34,
|
| 108 |
+
"MAC": 35,
|
| 109 |
+
"MASKEDNUMBER": 36,
|
| 110 |
+
"MIDDLENAME": 37,
|
| 111 |
+
"NEARBYGPSCOORDINATE": 38,
|
| 112 |
+
"O": 0,
|
| 113 |
+
"ORDINALDIRECTION": 39,
|
| 114 |
+
"PASSWORD": 40,
|
| 115 |
+
"PHONEIMEI": 41,
|
| 116 |
+
"PHONENUMBER": 42,
|
| 117 |
+
"PIN": 43,
|
| 118 |
+
"PREFIX": 44,
|
| 119 |
+
"SECONDARYADDRESS": 45,
|
| 120 |
+
"SEX": 46,
|
| 121 |
+
"SSN": 47,
|
| 122 |
+
"STATE": 48,
|
| 123 |
+
"STREET": 49,
|
| 124 |
+
"TIME": 50,
|
| 125 |
+
"URL": 51,
|
| 126 |
+
"USERAGENT": 52,
|
| 127 |
+
"USERNAME": 53,
|
| 128 |
+
"VEHICLEVIN": 54,
|
| 129 |
+
"VEHICLEVRM": 55,
|
| 130 |
+
"ZIPCODE": 56
|
| 131 |
+
},
|
| 132 |
+
"layer_norm_eps": 1e-07,
|
| 133 |
+
"legacy": true,
|
| 134 |
+
"max_position_embeddings": 512,
|
| 135 |
+
"max_relative_positions": -1,
|
| 136 |
+
"model_type": "deberta-v2",
|
| 137 |
+
"norm_rel_ebd": "layer_norm",
|
| 138 |
+
"num_attention_heads": 12,
|
| 139 |
+
"num_hidden_layers": 12,
|
| 140 |
+
"pad_token_id": 0,
|
| 141 |
+
"pooler_dropout": 0,
|
| 142 |
+
"pooler_hidden_act": "gelu",
|
| 143 |
+
"pooler_hidden_size": 768,
|
| 144 |
+
"pos_att_type": [
|
| 145 |
+
"p2c",
|
| 146 |
+
"c2p"
|
| 147 |
+
],
|
| 148 |
+
"position_biased_input": false,
|
| 149 |
+
"position_buckets": 256,
|
| 150 |
+
"relative_attention": true,
|
| 151 |
+
"share_att_key": true,
|
| 152 |
+
"tie_word_embeddings": true,
|
| 153 |
+
"transformers_version": "5.3.0",
|
| 154 |
+
"type_vocab_size": 0,
|
| 155 |
+
"use_cache": false,
|
| 156 |
+
"vocab_size": 128100
|
| 157 |
+
}
|
pii-masking-200k/SpanBased-CustomDeBERTa/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e05188643660894d8b832d6a08716f89c5e79ee08bd5c454e6fb1606ed79104b
|
| 3 |
+
size 741566223
|
pii-masking-200k/SpanBased-CustomDeBERTa/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pii-masking-200k/SpanBased-CustomDeBERTa/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 |
+
}
|