@mailwoman/neural-weights-fr-fr 9.2.0 → 9.4.0

package/README.md

@@ -80,7 +80,7 @@ metric on an anchor-trained model is systematically misleading. Construct the
 scorer through `@mailwoman/neural`'s `createScorer` (the canonical
 `ProductionScorer`), which reads the bundle's `requires`/channel contract and
 **fails closed** if a declared channel isn't fed. Do not hand-wire the raw ONNX
-session with the anchor input zero-filled. Every honest eval here is
+session with the anchor input zero-filled. Every eval here is
 **anchor-on**.
 
 ## Evaluation
@@ -92,7 +92,7 @@ neural front-end out-parses the rules baseline it replaces on locality-match.
 
 The two `model-index` entries above (postcode ≈ 0.99 F1, house_number ≈ 0.946 F1)
 are the multi-locale model's strong FR tags, **anchor-on, canonical order**. Two
-honest caveats on the FR numbers:
+known gaps on the FR numbers:
 
 - **FR `region` is a known open gap** — F1 is low and tracked as an outstanding
   issue; do not expect region-level accuracy on FR comparable to en-us.
@@ -100,8 +100,8 @@ honest caveats on the FR numbers:
   diversified golden that mixes both orders, house_number drops well below the
   canonical-order figure — the published frontier for reordered house numbers is
   ~90–91% (neural parsers collapse much further on reorder). The en-us release's
-  later checkpoint adds a reversed-order shard that recovers this; this fr-fr
-  bundle predates that lever.
+  later checkpoint adds a reversed-order training subset that recovers this; this fr-fr
+  bundle predates that change.
 
 Because the older anchor-off / pre-diversified-golden figures in this bundle's
 `model-card.json` `eval` block were measured before the anchor-on grading

package/model-card.json

@@ -1,12 +1,12 @@
 {
 	"name": "neural-weights-fr-fr",
 	"version": "6.5.0",
-	"model_lineage": "Ships the en-us v310 (v3.1.0-fr-fragment, #1143 T2) model — init_from v264 plus the synth-fr-fragment BAN shard @ 12.0; tokenizer v0.9.0-multisplice (73,143) UNCHANGED. One model serves both locales; see the en-us card for the full lineage + grade. NOTE for this locale specifically: the shard is BAN-derived (Base Adresse Nationale, Licence Ouverte 2.0) and the gains are measured on French — bare-street 0.215→0.715, street-particle 0.273→0.855, admin-street-homonym 0.087→0.517, with the bare-locality counter-guard holding exactly. This is the largest FR-facing model change since 5.9.0.",
+	"model_lineage": "Ships the en-us v310 (v3.1.0-fr-fragment, #1143 T2) model — init_from v264 plus the synth-fr-fragment BAN extract @ 12.0; tokenizer v0.9.0-multisplice (73,143) UNCHANGED. One model serves both locales; see the en-us card for the full lineage + grade. NOTE for this locale specifically: the extract is BAN-derived (Base Adresse Nationale, Licence Ouverte 2.0) and the gains are measured on French — bare-street 0.215→0.715, street-particle 0.273→0.855, admin-street-homonym 0.087→0.517, with the bare-locality counter-guard holding exactly. This is the largest FR-facing model change since 5.9.0.",
 	"phase": "Stage 3 — multi-locale (FR served by the shared en-us v310 model, shipped as 6.4.0)",
 	"license": "AGPL-3.0-only OR LicenseRef-Commercial",
 	"locale": "fr-fr",
 	"training": {
-		"corpus_version": "v257 fragment-campaign corpus (shard-v5) re-tokenized through v0.9.0-multisplice; span-boundary aux head from BIO labels.",
+		"corpus_version": "v257 fragment-campaign corpus (extract-v5) re-tokenized through v0.9.0-multisplice; span-boundary aux head from BIO labels.",
 		"tokenizer_version": "0.9.0-multisplice",
 		"steps": 8000,
 		"hardware": "NVIDIA A100-SXM4-40GB (Modal cloud)",
@@ -266,7 +266,7 @@
 		"tokenizer": "tokenizer.model",
 		"model_card": "model-card.json",
 		"pair_index": "pair-index-fr.bin",
-		"$comment_postcode_anchor": "Added 2026-08-06 (#1516), same reason as the en-us card: the overlay has always shipped this binary (release.config.json softFeed.postcodeDBByCountry.fr, built from the WOF intl postcode shard) and never declared it. `readDeclaredArtifactFile` reads this key to tell a broken package (declared, missing) from a package that deliberately ships none.",
+		"$comment_postcode_anchor": "Added 2026-08-06 (#1516), same reason as the en-us card: the overlay has always shipped this binary (release.config.json softFeed.postcodeDBByCountry.fr, built from the WOF intl postcode extract) and never declared it. `readDeclaredArtifactFile` reads this key to tell a broken package (declared, missing) from a package that deliberately ships none.",
 		"postcode_anchor": "postcode-fr.bin"
 	},
 	"base_relpath": "/data/models/checkpoints/v0_4_0-stableLR-source-only/step-002200",
@@ -280,7 +280,7 @@
 			"md5": "(staged at release time — see $comment_md5)",
 			"$comment_pix2": "PIX2 (schema 3, 2026-08-04): every record now carries the PARENT's ComponentTag alongside the child's — +1 byte per pair, pair COUNT unchanged. For FR the recorded parent tag is always `locality`: BAN's parent surface is `nom_commune`, the commune, which is the French postal locality line. Schema 1 and 2 binaries are refused; the artifact must be rebuilt.",
 			"delta_calibration": "δ=10, transitionBeta=5, parentDelta=5 (whole-edge parent bias, #46 — the record's own parentTag gets +5 over the parent window). FR is the leg that CAUGHT that mechanism's one real defect: biasing the parent's whole comma segment put the postcode inside `locality` and took fr-lieudit-golden.jsonl from 96.3% to 0.0% at δ≥6, invisible at δ=4. Fixed by scoping the parent write to the probe key's pieces; FR then held 77/80 at every δ through 20. See docs/records/evals/2026-08-04-pix1-whole-edge-verdict.md. The δ/β pair is inherited from the GB calibration, per R5's finding that the classifier's dependent_locality deficit is uniform across locales rather than GB-specific. The R6 bars were measured at exactly that pair.",
-			"country_gate": "hard-gated on the header's country (fr).",
+			"country_check": "hard-blocked on the header's country (fr).",
 			"not_committed": "derived artifact, .gitignore'd — rebuilt via `mailwoman gazetteer pair-index --out neural-weights-fr-fr --country fr --delta 10 --transition-beta 5 --ban-dir <corpus/sources/ban>`",
 			"required_decode_fixes": "This artifact is INERT without two locale-general probe-chain fixes that shipped alongside it: a LEADING postcode strip on parent-candidate keys (France writes '12210 Montpeyroux'; the probe previously stripped only trailing postcodes, the Anglo/NZ convention) and NEWLINE as a segment boundary (computeGroupSegments scanned for commas only, so multi-line addresses collapsed to one segment and the segment path went structurally inert). With the artifact but without them the golden board stayed at 0/80.",
 			"$comment_md5": "REPRODUCIBILITY: a PIX1 header embeds `buildDate`, so a rebuild from identical sources produces different bytes and a different md5. This md5 therefore identifies the artifact STAGED TO HUGGING FACE for this release — it is a provenance record of the shipped bytes, NOT a check a local rebuild can reproduce. A mismatch after rebuilding is expected and is not corruption; verify pair COUNT and header delta/beta instead.",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@mailwoman/neural-weights-fr-fr",
-  "version": "9.2.0",
+  "version": "9.4.0",
   "description": "Mailwoman neural-classifier weights for locale 'fr-fr'. Data-only overlay — shares the base model.onnx + tokenizer.model from @mailwoman/neural-weights-en-us (they are byte-identical); ships only the fr-specific data siblings. Loaded by @mailwoman/neural at runtime.",
   "license": "AGPL-3.0-only OR LicenseRef-Commercial",
   "repository": {
@@ -30,6 +30,13 @@
     "!scripts/**",
     "!test/**"
   ],
+  "imports": {
+    "#*": {
+      "types": "./out/*.d.ts",
+      "node": "./out/*.js",
+      "default": "./out/*.js"
+    }
+  },
   "exports": {
     "./*": "./*"
   },
@@ -37,14 +44,21 @@
     "access": "public",
     "exports": {
       "./*": "./*"
+    },
+    "imports": {
+      "#*": {
+        "types": "./out/*.d.ts",
+        "node": "./out/*.js",
+        "default": "./out/*.js"
+      }
     }
   },
   "dependencies": {
-    "@mailwoman/neural-weights-en-us": "9.2.0"
+    "@mailwoman/neural-weights-en-us": "9.4.0"
   },
   "devDependencies": {
-    "@mailwoman/core": "9.2.0",
-    "@mailwoman/resolver-wof-sqlite": "9.2.0"
+    "@mailwoman/core": "9.4.0",
+    "@mailwoman/resolver-wof-sqlite": "9.4.0"
   },
   "mailwoman": {
     "baseWeights": "@mailwoman/neural-weights-en-us"