Upload ModernBERT PII NER model

README.md

@@ -0,0 +1,50 @@
+---
+license: apache-2.0
+base_model: answerdotai/ModernBERT-base
+tags:
+  - token-classification
+  - named-entity-recognition
+  - pii
+  - modernbert
+language:
+  - en
+pipeline_tag: token-classification
+---
+
+# modernbert-pii
+
+ModernBERT fine-tuned for token classification (NER) to detect PII in
+Australian clinical notes. Labels use the BIO scheme over these entity types:
+
+- `ACD`
+- `ADDRESS`
+- `AGE`
+- `ATTENDING_NURSE`
+- `COUNTRY_OF_BIRTH`
+- `DATE_OF_BIRTH`
+- `DVA_NUMBER`
+- `EMAIL`
+- `FULL_NAME`
+- `GUARDIANSHIP`
+- `IHI`
+- `MEDICARE_NUMBER`
+- `MRN`
+- `NEXT_OF_KIN`
+- `OCCUPATION`
+- `OTHER_CLINICAL_STAFF`
+- `PHONE_NUMBER`
+- `PREFERRED_NAME`
+- `PRIVATE_HEALTH_NUMBER`
+- `REFERRING_GP`
+- `SEX_GENDER`
+- `TREATING_CLINICIAN`
+- `WORKCOVER`
+
+## Usage
+
+```python
+from transformers import pipeline
+
+ner = pipeline("token-classification", model="Cbezz/modernbert-pii", aggregation_strategy="simple")
+ner("Patient John Smith, DOB 12/03/1980, MRN 4456219.")
+```

config.json

@@ -0,0 +1,176 @@
+{
+  "architectures": [
+    "ModernBertForTokenClassification"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": null,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "dtype": "float32",
+  "embedding_dropout": 0.0,
+  "eos_token_id": null,
+  "global_attn_every_n_layers": 3,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "id2label": {
+    "0": "O",
+    "1": "B-ACD",
+    "2": "I-ACD",
+    "3": "B-ADDRESS",
+    "4": "I-ADDRESS",
+    "5": "B-AGE",
+    "6": "I-AGE",
+    "7": "B-ATTENDING_NURSE",
+    "8": "I-ATTENDING_NURSE",
+    "9": "B-COUNTRY_OF_BIRTH",
+    "10": "I-COUNTRY_OF_BIRTH",
+    "11": "B-DATE_OF_BIRTH",
+    "12": "I-DATE_OF_BIRTH",
+    "13": "B-DVA_NUMBER",
+    "14": "I-DVA_NUMBER",
+    "15": "B-EMAIL",
+    "16": "I-EMAIL",
+    "17": "B-FULL_NAME",
+    "18": "I-FULL_NAME",
+    "19": "B-GUARDIANSHIP",
+    "20": "I-GUARDIANSHIP",
+    "21": "B-IHI",
+    "22": "I-IHI",
+    "23": "B-MEDICARE_NUMBER",
+    "24": "I-MEDICARE_NUMBER",
+    "25": "B-MRN",
+    "26": "I-MRN",
+    "27": "B-NEXT_OF_KIN",
+    "28": "I-NEXT_OF_KIN",
+    "29": "B-OCCUPATION",
+    "30": "I-OCCUPATION",
+    "31": "B-OTHER_CLINICAL_STAFF",
+    "32": "I-OTHER_CLINICAL_STAFF",
+    "33": "B-PHONE_NUMBER",
+    "34": "I-PHONE_NUMBER",
+    "35": "B-PREFERRED_NAME",
+    "36": "I-PREFERRED_NAME",
+    "37": "B-PRIVATE_HEALTH_NUMBER",
+    "38": "I-PRIVATE_HEALTH_NUMBER",
+    "39": "B-REFERRING_GP",
+    "40": "I-REFERRING_GP",
+    "41": "B-SEX_GENDER",
+    "42": "I-SEX_GENDER",
+    "43": "B-TREATING_CLINICIAN",
+    "44": "I-TREATING_CLINICIAN",
+    "45": "B-WORKCOVER",
+    "46": "I-WORKCOVER"
+  },
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "label2id": {
+    "B-ACD": 1,
+    "B-ADDRESS": 3,
+    "B-AGE": 5,
+    "B-ATTENDING_NURSE": 7,
+    "B-COUNTRY_OF_BIRTH": 9,
+    "B-DATE_OF_BIRTH": 11,
+    "B-DVA_NUMBER": 13,
+    "B-EMAIL": 15,
+    "B-FULL_NAME": 17,
+    "B-GUARDIANSHIP": 19,
+    "B-IHI": 21,
+    "B-MEDICARE_NUMBER": 23,
+    "B-MRN": 25,
+    "B-NEXT_OF_KIN": 27,
+    "B-OCCUPATION": 29,
+    "B-OTHER_CLINICAL_STAFF": 31,
+    "B-PHONE_NUMBER": 33,
+    "B-PREFERRED_NAME": 35,
+    "B-PRIVATE_HEALTH_NUMBER": 37,
+    "B-REFERRING_GP": 39,
+    "B-SEX_GENDER": 41,
+    "B-TREATING_CLINICIAN": 43,
+    "B-WORKCOVER": 45,
+    "I-ACD": 2,
+    "I-ADDRESS": 4,
+    "I-AGE": 6,
+    "I-ATTENDING_NURSE": 8,
+    "I-COUNTRY_OF_BIRTH": 10,
+    "I-DATE_OF_BIRTH": 12,
+    "I-DVA_NUMBER": 14,
+    "I-EMAIL": 16,
+    "I-FULL_NAME": 18,
+    "I-GUARDIANSHIP": 20,
+    "I-IHI": 22,
+    "I-MEDICARE_NUMBER": 24,
+    "I-MRN": 26,
+    "I-NEXT_OF_KIN": 28,
+    "I-OCCUPATION": 30,
+    "I-OTHER_CLINICAL_STAFF": 32,
+    "I-PHONE_NUMBER": 34,
+    "I-PREFERRED_NAME": 36,
+    "I-PRIVATE_HEALTH_NUMBER": 38,
+    "I-REFERRING_GP": 40,
+    "I-SEX_GENDER": 42,
+    "I-TREATING_CLINICIAN": 44,
+    "I-WORKCOVER": 46,
+    "O": 0
+  },
+  "layer_norm_eps": 1e-05,
+  "layer_types": [
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "local_attention": 128,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "rope_parameters": {
+    "full_attention": {
+      "rope_theta": 160000.0,
+      "rope_type": "default"
+    },
+    "sliding_attention": {
+      "rope_theta": 10000.0,
+      "rope_type": "default"
+    }
+  },
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.3.0",
+  "use_cache": false,
+  "vocab_size": 50368
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:415a47eb4646e8cf83e2a2e3c9135ce712e763dd122eff6c394d1fa82a9f3e25
+size 598578212

tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "backend": "tokenizers",
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "is_local": false,
+  "mask_token": "[MASK]",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 8192,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": "[UNK]"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11f76064b6b389e75db95abcb272d01b3190bff761ce32639bb6df48b5939637
+size 4792