Model card: link the canonical leaderboard; declare the fine-tune relation to the base model

README.md

@@ -22,7 +22,7 @@ model-index:
       value: 0.7766
       name: F1, character-level, label-agnostic, averaged over 7 datasets
     - type: precision
-      value: 0.7100
+      value: 0.7104
       name: Precision
     - type: recall
       value: 0.8780
@@ -78,8 +78,12 @@ each dataset (1,000 resamples for each model's F1; 2,000 for the paired differen
 | precision | 0.710 | 0.744 |
 | recall | **0.878** | 0.783 |
 | touch rate (not ranked; lower is better) | **0.021** | 0.288 |
-| latency, median ms per row, batch size 1 | **20.3** | 140.6 |
-| latency, 95th percentile ms per row | **33.5** | 199.7 |
+| latency, median ms per row, batch size 1 | **20.3** † | 140.6 |
+| latency, 95th percentile ms per row | **33.5** † | 199.7 |
+
+† Measured on the previous release (v15, revision `e98225e07225`), which has the same
+architecture and is exported to int8 ONNX the same way; latency has not been re-measured for
+this build.
 
 The paired difference in average F1 is **+0.0233, 95 % CI [+0.0147, +0.0322]** (2,000
 resamples, seed 0); the interval excludes zero.
@@ -90,8 +94,8 @@ Three qualifications belong with that result:
   entity leaves the device unredacted while a false positive masks a word unnecessarily, so
   the trade is in the intended direction — but it is a trade, and the F1 average conceals it.
 - **Latency was measured on a shared workstation** under a higher load average for this model
-  (1-minute load 20.7) than for the comparison model (11.1). Absolute milliseconds depend on the
-  host; the ordering is the informative part.
+  (1-minute load 20.7) than for the comparison model (11.1), and for this model on the previous
+  release (†). Absolute milliseconds depend on the host; the ordering is the informative part.
 - **Touch rate** is the share of characters belonging to format-fixed identifiers that a model
   marks anyway. It is reported, not ranked.