add reference to gliner2 zero shot capabilities

README.md

@@ -84,6 +84,8 @@ The synthetic data approach effectively distils the "knowledge" of a large LLM i
 | `TECHNICAL_ID_NUMBERS` | IP/MAC addresses, serial numbers |
 | `VEHICLE_ID_NUMBERS` | License plates, VINs |
 
+Since NERPA is built on GLiNER2 (a zero-shot bi-encoder), it is **not limited** to the entities above. You can pass any custom entity types alongside the built-in ones — the fine-tuning does not reduce the model's ability to detect arbitrary categories. See [Custom entities](#custom-entities) below.
+
 ## Quick Start
 
 ### Install dependencies
@@ -158,6 +160,41 @@ entities = detect_entities(model, text, entities={
 })
 ```
 
+### Custom entities
+
+You can detect additional entity types beyond the built-in PII set. The model's zero-shot capability means any label + description pair will work — your custom entities are detected and anonymised alongside the fine-tuned ones.
+
+**CLI** — use `--extra-entities` / `-e`:
+
+```bash
+python anonymise.py -e PRODUCT="Product name" -e SKILL="Professional skill" \
+    "John Smith is a senior Python developer who bought a MacBook Pro."
+```
+
+Output:
+
+```
+[PERSON_NAME] is a senior [SKILL] developer who bought a [PRODUCT].
+```
+
+**Python:**
+
+```python
+from anonymise import load_model, detect_entities, anonymise, PII_ENTITIES
+
+model = load_model(".")
+
+custom_entities = {
+    **PII_ENTITIES,
+    "PRODUCT": "Product name",
+    "SKILL": "Professional skill",
+}
+
+text = "John Smith is a senior Python developer who bought a MacBook Pro."
+entities = detect_entities(model, text, entities=custom_entities)
+print(anonymise(text, entities))
+```
+
 ## How It Works
 
 The inference pipeline in `anonymise.py`:

anonymise.py

@@ -212,6 +212,14 @@ def main() -> None:
         "--show-entities", action="store_true",
         help="Print detected entities before anonymised text",
     )
+    parser.add_argument(
+        "--extra-entities", "-e", action="append", metavar="LABEL=DESCRIPTION",
+        help=(
+            "Additional custom entity types to detect alongside the built-in "
+            "PII entities. Repeat for each type.  Format: LABEL=\"Description\".  "
+            "Example: -e PRODUCT=\"Product name\" -e SKILL=\"Professional skill\""
+        ),
+    )
     args = parser.parse_args()
 
     if args.file:
@@ -225,8 +233,20 @@ def main() -> None:
     else:
         parser.error("Provide text as an argument or use --file")
 
+    extra: dict[str, str] = {}
+    if args.extra_entities:
+        for item in args.extra_entities:
+            if "=" not in item:
+                parser.error(
+                    f"Invalid --extra-entities value '{item}'. "
+                    "Expected format: LABEL=\"Description\""
+                )
+            label, description = item.split("=", 1)
+            extra[label.strip()] = description.strip()
+
     model = load_model(args.model)
-    detected = detect_entities(model, text, threshold=args.threshold)
+    all_entities = {**PII_ENTITIES, **extra} if extra else None
+    detected = detect_entities(model, text, entities=all_entities, threshold=args.threshold)
 
     if args.show_entities:
         for entity in detected: