chore: upgrade data-designer to 0.6.0

Makefile

@@ -129,7 +129,7 @@ convert-notebooks:
 	@echo "Converting Python tutorials to notebooks and executing..."
 	@mkdir -p docs/notebooks
 	uv run --group notebooks python -m ipykernel install --user --name anonymizer-venv
-	uv run --group notebooks --group docs jupytext --to ipynb --execute docs/notebook_source/*.py
+	uv run --group notebooks --group docs jupytext --to ipynb --set-kernel anonymizer-venv --execute docs/notebook_source/*.py
 	mv docs/notebook_source/*.ipynb docs/notebooks/
 	@echo "Notebooks created in docs/notebooks/"
 

docs/notebook_source/04_rewriting_biographies.py

@@ -57,8 +57,7 @@
     os.environ["NVIDIA_API_KEY"] = key
 
 # %%
-from anonymizer import Anonymizer, AnonymizerConfig, AnonymizerInput, Rewrite, configure_logging
-from anonymizer.config.rewrite import PrivacyGoal
+from anonymizer import Anonymizer, AnonymizerConfig, AnonymizerInput, PrivacyGoal, Rewrite, configure_logging
 
 configure_logging(enabled=False)
 

docs/notebook_source/05_rewriting_legal_documents.py

@@ -52,8 +52,7 @@
     os.environ["NVIDIA_API_KEY"] = key
 
 # %%
-from anonymizer import Anonymizer, AnonymizerConfig, AnonymizerInput, Detect, Rewrite, configure_logging
-from anonymizer.config.rewrite import PrivacyGoal
+from anonymizer import Anonymizer, AnonymizerConfig, AnonymizerInput, Detect, PrivacyGoal, Rewrite, configure_logging
 
 configure_logging(enabled=False)
 

pyproject.toml

@@ -7,7 +7,7 @@ requires-python = ">=3.11"
 license = "Apache-2.0"
 
 dependencies = [
-    "data-designer==0.5.7",
+    "data-designer==0.6.0",
     "pydantic>=2.9,<3",
     "cyclopts>=3",
     "pygments>=2.20.0",

src/anonymizer/config/default_model_configs/models.yaml

@@ -13,7 +13,7 @@ model_configs:
     model: openai/gpt-oss-120b
     provider: nvidia
     inference_parameters:
-      max_parallel_requests: 2
+      max_parallel_requests: 16
       max_tokens: 16384
       temperature: 0.3
       top_p: 0.95

src/anonymizer/engine/replace/llm_replace_workflow.py

@@ -5,6 +5,7 @@
 
 import json
 import logging
+from collections import Counter
 from dataclasses import dataclass
 
 import pandas as pd
@@ -169,19 +170,27 @@ def _filter_replacement_map_to_input_entities(
         filtered.append(replacement.model_dump())
 
     if logger.isEnabledFor(logging.DEBUG):
+        raw_pairs = {(r.original, r.label) for r in parsed_map.replacements}
+        filtered_pairs = {(f["original"], f["label"]) for f in filtered}
+        unrequested_labels = Counter(label for _, label in (raw_pairs - allowed_pairs))
+        unfilled_labels = Counter(label for _, label in (allowed_pairs - filtered_pairs))
         logger.debug(
-            "Replacement map record %s: requested=%s raw=%s filtered=%s",
+            "Replacement map record %s: requested=%d raw=%d filtered=%d%s%s",
             record_id or "<unknown>",
-            sorted(allowed_pairs),
-            [entry.model_dump() for entry in parsed_map.replacements],
-            filtered,
+            len(allowed_pairs),
+            len(parsed_map.replacements),
+            len(filtered),
+            f" unrequested_by_label={dict(unrequested_labels)}" if unrequested_labels else "",
+            f" unfilled_by_label={dict(unfilled_labels)}" if unfilled_labels else "",
         )
-    elif not filtered and allowed_pairs:
+    if not filtered and allowed_pairs:
+        requested_labels = Counter(label for _, label in allowed_pairs)
         logger.warning(
-            "Replacement map empty after filtering for record %s; requested=%s raw=%s",
+            "Replacement map empty after filtering for record %s; requested=%d raw=%d (requested_by_label=%s)",
             record_id or "<unknown>",
-            sorted(allowed_pairs),
-            [entry.model_dump() for entry in parsed_map.replacements],
+            len(allowed_pairs),
+            len(parsed_map.replacements),
+            dict(requested_labels),
         )
     return {"replacements": filtered}
 

tests/engine/test_llm_replace_workflow.py

@@ -3,15 +3,22 @@
 
 from __future__ import annotations
 
+import logging
 from unittest.mock import Mock
 
 import pandas as pd
+import pytest
 from data_designer.config.models import ModelConfig
 
 from anonymizer.config.models import ReplaceModelSelection
 from anonymizer.engine.constants import COL_ENTITIES_BY_VALUE, COL_REPLACEMENT_MAP, COL_TEXT
 from anonymizer.engine.ndd.adapter import WorkflowRunResult
-from anonymizer.engine.replace.llm_replace_workflow import _INTERNAL_COLUMNS, LlmReplaceWorkflow
+from anonymizer.engine.replace.llm_replace_workflow import (
+    _INTERNAL_COLUMNS,
+    LlmReplaceWorkflow,
+    _filter_replacement_map_to_input_entities,
+)
+from anonymizer.engine.schemas.detection import EntitiesByValueSchema
 
 
 def test_generate_map_only_returns_replacement_map(
@@ -242,3 +249,110 @@ def test_generate_map_only_strips_internal_prompt_columns_when_no_entities(
     adapter.run_workflow.assert_not_called()
     for col in _INTERNAL_COLUMNS:
         assert col not in result.dataframe.columns, f"workflow-internal column {col!r} leaked into result"
+
+
+# ---------------------------------------------------------------------------
+# PII-free logging regression tests for _filter_replacement_map_to_input_entities
+#
+# The replacement map filter runs after the LLM proposes substitutions and
+# emits a DEBUG summary plus a WARNING when the filter empties out. Both
+# log paths must report counts and labels only, never raw entity values.
+# ---------------------------------------------------------------------------
+
+_ORIGINAL_PII = ("Jane Doe", "jane.doe@example.com", "+1-555-867-5309")
+_SYNTHETIC_PII = ("Maya Chen", "maya.chen@example.com", "+1-555-000-1234")
+
+
+def _assert_no_pii_in_logs(caplog: pytest.LogCaptureFixture, extra_secrets: tuple[str, ...] = ()) -> None:
+    for secret in (*_ORIGINAL_PII, *_SYNTHETIC_PII, *extra_secrets):
+        assert secret not in caplog.text, f"PII leak in logs: {secret!r} appeared in:\n{caplog.text}"
+
+
+def test_filter_replacement_map_debug_log_does_not_leak_pii(
+    caplog: pytest.LogCaptureFixture,
+) -> None:
+    """The DEBUG summary on the happy path must not emit raw entity values."""
+    parsed_entities = EntitiesByValueSchema.model_validate(
+        {
+            "entities_by_value": [
+                {"value": "Jane Doe", "labels": ["first_name"]},
+                {"value": "jane.doe@example.com", "labels": ["email"]},
+                {"value": "+1-555-867-5309", "labels": ["phone_number"]},
+            ]
+        }
+    )
+    raw_map = {
+        "replacements": [
+            {"original": "Jane Doe", "label": "first_name", "synthetic": "Maya Chen"},
+            {"original": "jane.doe@example.com", "label": "email", "synthetic": "maya.chen@example.com"},
+            {"original": "+1-555-867-5309", "label": "phone_number", "synthetic": "+1-555-000-1234"},
+        ]
+    }
+
+    with caplog.at_level(logging.DEBUG, logger="anonymizer"):
+        result = _filter_replacement_map_to_input_entities(
+            raw_map=raw_map, parsed_entities=parsed_entities, record_id="row-abc123"
+        )
+
+    assert "Replacement map record" in caplog.text
+    assert "requested=3" in caplog.text
+    _assert_no_pii_in_logs(caplog)
+    assert len(result["replacements"]) == 3
+
+
+def test_filter_replacement_map_anomaly_summaries_do_not_leak_pii(
+    caplog: pytest.LogCaptureFixture,
+) -> None:
+    """Unrequested-by-label / unfilled-by-label extras must name the LABEL, not the value."""
+    parsed_entities = EntitiesByValueSchema.model_validate(
+        {
+            "entities_by_value": [
+                {"value": "Jane Doe", "labels": ["first_name"]},
+                {"value": "+1-555-867-5309", "labels": ["phone_number"]},
+            ]
+        }
+    )
+    raw_map = {
+        "replacements": [
+            {"original": "Jane Doe", "label": "first_name", "synthetic": "Maya Chen"},
+            {"original": "Acme Corp", "label": "organization_name", "synthetic": "NovaCorp"},
+        ]
+    }
+
+    with caplog.at_level(logging.DEBUG, logger="anonymizer"):
+        _filter_replacement_map_to_input_entities(raw_map=raw_map, parsed_entities=parsed_entities, record_id="row-xyz")
+
+    assert "unrequested_by_label" in caplog.text
+    assert "unfilled_by_label" in caplog.text
+    assert "organization_name" in caplog.text
+    assert "phone_number" in caplog.text
+
+    _assert_no_pii_in_logs(caplog, extra_secrets=("Acme Corp", "NovaCorp"))
+
+
+def test_filter_replacement_map_empty_warning_does_not_leak_pii(
+    caplog: pytest.LogCaptureFixture,
+) -> None:
+    """The empty-after-filtering WARNING must report counts + labels only."""
+    parsed_entities = EntitiesByValueSchema.model_validate(
+        {
+            "entities_by_value": [
+                {"value": "Jane Doe", "labels": ["first_name"]},
+            ]
+        }
+    )
+    raw_map = {
+        "replacements": [
+            {"original": "Acme Corp", "label": "organization_name", "synthetic": "NovaCorp"},
+        ]
+    }
+
+    with caplog.at_level(logging.WARNING, logger="anonymizer"):
+        result = _filter_replacement_map_to_input_entities(
+            raw_map=raw_map, parsed_entities=parsed_entities, record_id="row-empty"
+        )
+
+    assert "Replacement map empty after filtering" in caplog.text
+    assert "first_name" in caplog.text
+    _assert_no_pii_in_logs(caplog, extra_secrets=("Acme Corp", "NovaCorp"))
+    assert result == {"replacements": []}