Adding ONNX file of this model (#1)

- Adding ONNX file of this model (9dc48efd31ba5c6687ad0e5479daf8abe7f66293)

onnx/added_tokens.json

@@ -0,0 +1,7 @@
+{
+  "[CLS]": 101,
+  "[MASK]": 103,
+  "[PAD]": 0,
+  "[SEP]": 102,
+  "[UNK]": 100
+}

onnx/config.json

@@ -0,0 +1,249 @@
+{
+  "_name_or_path": "Isotonic/distilbert_finetuned_ai4privacy_v2",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForTokenClassification"
+  ],
+  "attention_dropout": 0.1,
+  "dim": 768,
+  "dropout": 0.1,
+  "hidden_dim": 3072,
+  "id2label": {
+    "0": "O",
+    "1": "B-PHONEIMEI",
+    "2": "I-PHONEIMEI",
+    "3": "B-JOBAREA",
+    "4": "B-FIRSTNAME",
+    "5": "I-FIRSTNAME",
+    "6": "B-VEHICLEVIN",
+    "7": "I-VEHICLEVIN",
+    "8": "B-AGE",
+    "9": "B-GENDER",
+    "10": "I-GENDER",
+    "11": "B-HEIGHT",
+    "12": "I-HEIGHT",
+    "13": "B-BUILDINGNUMBER",
+    "14": "I-BUILDINGNUMBER",
+    "15": "B-MASKEDNUMBER",
+    "16": "I-MASKEDNUMBER",
+    "17": "B-PASSWORD",
+    "18": "I-PASSWORD",
+    "19": "B-DOB",
+    "20": "I-DOB",
+    "21": "B-IPV6",
+    "22": "I-IPV6",
+    "23": "B-NEARBYGPSCOORDINATE",
+    "24": "I-NEARBYGPSCOORDINATE",
+    "25": "B-USERAGENT",
+    "26": "I-USERAGENT",
+    "27": "B-TIME",
+    "28": "I-TIME",
+    "29": "B-JOBTITLE",
+    "30": "I-JOBTITLE",
+    "31": "B-COUNTY",
+    "32": "B-EMAIL",
+    "33": "I-EMAIL",
+    "34": "B-ACCOUNTNUMBER",
+    "35": "I-ACCOUNTNUMBER",
+    "36": "B-PIN",
+    "37": "I-PIN",
+    "38": "B-EYECOLOR",
+    "39": "I-EYECOLOR",
+    "40": "B-LASTNAME",
+    "41": "I-LASTNAME",
+    "42": "I-JOBAREA",
+    "43": "B-IPV4",
+    "44": "I-IPV4",
+    "45": "B-DATE",
+    "46": "I-DATE",
+    "47": "B-STREET",
+    "48": "I-STREET",
+    "49": "B-CITY",
+    "50": "I-CITY",
+    "51": "B-PREFIX",
+    "52": "I-PREFIX",
+    "53": "B-CREDITCARDISSUER",
+    "54": "B-CREDITCARDNUMBER",
+    "55": "I-CREDITCARDNUMBER",
+    "56": "I-CREDITCARDISSUER",
+    "57": "B-MIDDLENAME",
+    "58": "B-STATE",
+    "59": "I-STATE",
+    "60": "B-VEHICLEVRM",
+    "61": "I-VEHICLEVRM",
+    "62": "B-ORDINALDIRECTION",
+    "63": "B-SEX",
+    "64": "B-JOBTYPE",
+    "65": "I-JOBTYPE",
+    "66": "B-CURRENCYCODE",
+    "67": "I-CURRENCYCODE",
+    "68": "B-CURRENCYSYMBOL",
+    "69": "I-AMOUNT",
+    "70": "B-ACCOUNTNAME",
+    "71": "I-ACCOUNTNAME",
+    "72": "B-BITCOINADDRESS",
+    "73": "I-BITCOINADDRESS",
+    "74": "B-LITECOINADDRESS",
+    "75": "I-LITECOINADDRESS",
+    "76": "B-PHONENUMBER",
+    "77": "I-PHONENUMBER",
+    "78": "B-MAC",
+    "79": "I-MAC",
+    "80": "B-CURRENCY",
+    "81": "B-IBAN",
+    "82": "I-IBAN",
+    "83": "B-COMPANYNAME",
+    "84": "I-COMPANYNAME",
+    "85": "B-CURRENCYNAME",
+    "86": "I-CURRENCYNAME",
+    "87": "I-CURRENCYSYMBOL",
+    "88": "B-ZIPCODE",
+    "89": "I-ZIPCODE",
+    "90": "B-SSN",
+    "91": "I-SSN",
+    "92": "B-AMOUNT",
+    "93": "I-CURRENCY",
+    "94": "B-URL",
+    "95": "I-URL",
+    "96": "B-IP",
+    "97": "I-IP",
+    "98": "B-SECONDARYADDRESS",
+    "99": "I-SECONDARYADDRESS",
+    "100": "B-USERNAME",
+    "101": "I-USERNAME",
+    "102": "B-ETHEREUMADDRESS",
+    "103": "I-ETHEREUMADDRESS",
+    "104": "B-CREDITCARDCVV",
+    "105": "I-CREDITCARDCVV",
+    "106": "I-COUNTY",
+    "107": "I-AGE",
+    "108": "I-MIDDLENAME",
+    "109": "B-BIC",
+    "110": "I-BIC"
+  },
+  "initializer_range": 0.02,
+  "label2id": {
+    "B-ACCOUNTNAME": 70,
+    "B-ACCOUNTNUMBER": 34,
+    "B-AGE": 8,
+    "B-AMOUNT": 92,
+    "B-BIC": 109,
+    "B-BITCOINADDRESS": 72,
+    "B-BUILDINGNUMBER": 13,
+    "B-CITY": 49,
+    "B-COMPANYNAME": 83,
+    "B-COUNTY": 31,
+    "B-CREDITCARDCVV": 104,
+    "B-CREDITCARDISSUER": 53,
+    "B-CREDITCARDNUMBER": 54,
+    "B-CURRENCY": 80,
+    "B-CURRENCYCODE": 66,
+    "B-CURRENCYNAME": 85,
+    "B-CURRENCYSYMBOL": 68,
+    "B-DATE": 45,
+    "B-DOB": 19,
+    "B-EMAIL": 32,
+    "B-ETHEREUMADDRESS": 102,
+    "B-EYECOLOR": 38,
+    "B-FIRSTNAME": 4,
+    "B-GENDER": 9,
+    "B-HEIGHT": 11,
+    "B-IBAN": 81,
+    "B-IP": 96,
+    "B-IPV4": 43,
+    "B-IPV6": 21,
+    "B-JOBAREA": 3,
+    "B-JOBTITLE": 29,
+    "B-JOBTYPE": 64,
+    "B-LASTNAME": 40,
+    "B-LITECOINADDRESS": 74,
+    "B-MAC": 78,
+    "B-MASKEDNUMBER": 15,
+    "B-MIDDLENAME": 57,
+    "B-NEARBYGPSCOORDINATE": 23,
+    "B-ORDINALDIRECTION": 62,
+    "B-PASSWORD": 17,
+    "B-PHONEIMEI": 1,
+    "B-PHONENUMBER": 76,
+    "B-PIN": 36,
+    "B-PREFIX": 51,
+    "B-SECONDARYADDRESS": 98,
+    "B-SEX": 63,
+    "B-SSN": 90,
+    "B-STATE": 58,
+    "B-STREET": 47,
+    "B-TIME": 27,
+    "B-URL": 94,
+    "B-USERAGENT": 25,
+    "B-USERNAME": 100,
+    "B-VEHICLEVIN": 6,
+    "B-VEHICLEVRM": 60,
+    "B-ZIPCODE": 88,
+    "I-ACCOUNTNAME": 71,
+    "I-ACCOUNTNUMBER": 35,
+    "I-AGE": 107,
+    "I-AMOUNT": 69,
+    "I-BIC": 110,
+    "I-BITCOINADDRESS": 73,
+    "I-BUILDINGNUMBER": 14,
+    "I-CITY": 50,
+    "I-COMPANYNAME": 84,
+    "I-COUNTY": 106,
+    "I-CREDITCARDCVV": 105,
+    "I-CREDITCARDISSUER": 56,
+    "I-CREDITCARDNUMBER": 55,
+    "I-CURRENCY": 93,
+    "I-CURRENCYCODE": 67,
+    "I-CURRENCYNAME": 86,
+    "I-CURRENCYSYMBOL": 87,
+    "I-DATE": 46,
+    "I-DOB": 20,
+    "I-EMAIL": 33,
+    "I-ETHEREUMADDRESS": 103,
+    "I-EYECOLOR": 39,
+    "I-FIRSTNAME": 5,
+    "I-GENDER": 10,
+    "I-HEIGHT": 12,
+    "I-IBAN": 82,
+    "I-IP": 97,
+    "I-IPV4": 44,
+    "I-IPV6": 22,
+    "I-JOBAREA": 42,
+    "I-JOBTITLE": 30,
+    "I-JOBTYPE": 65,
+    "I-LASTNAME": 41,
+    "I-LITECOINADDRESS": 75,
+    "I-MAC": 79,
+    "I-MASKEDNUMBER": 16,
+    "I-MIDDLENAME": 108,
+    "I-NEARBYGPSCOORDINATE": 24,
+    "I-PASSWORD": 18,
+    "I-PHONEIMEI": 2,
+    "I-PHONENUMBER": 77,
+    "I-PIN": 37,
+    "I-PREFIX": 52,
+    "I-SECONDARYADDRESS": 99,
+    "I-SSN": 91,
+    "I-STATE": 59,
+    "I-STREET": 48,
+    "I-TIME": 28,
+    "I-URL": 95,
+    "I-USERAGENT": 26,
+    "I-USERNAME": 101,
+    "I-VEHICLEVIN": 7,
+    "I-VEHICLEVRM": 61,
+    "I-ZIPCODE": 89,
+    "O": 0
+  },
+  "max_position_embeddings": 512,
+  "model_type": "distilbert",
+  "n_heads": 12,
+  "n_layers": 6,
+  "pad_token_id": 0,
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "transformers_version": "4.34.0",
+  "vocab_size": 30522
+}

onnx/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:415bcff5ce832675823ad56fc36082caf4f10c50673f30c2a43f0ed5fcd4a432
+size 265928108

onnx/special_tokens_map.json

@@ -0,0 +1,14 @@
+{
+  "additional_special_tokens": [
+    "[PAD]",
+    "[UNK]",
+    "[CLS]",
+    "[SEP]",
+    "[MASK]"
+  ],
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

onnx/tokenizer_config.json

@@ -0,0 +1,69 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "[PAD]",
+    "[UNK]",
+    "[CLS]",
+    "[SEP]",
+    "[MASK]"
+  ],
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "max_length": 512,
+  "model_max_length": 512,
+  "pad_to_multiple_of": null,
+  "pad_token": "[PAD]",
+  "pad_token_type_id": 0,
+  "padding_side": "right",
+  "sep_token": "[SEP]",
+  "stride": 0,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "DistilBertTokenizer",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": "[UNK]"
+}