Document observed specialization and cached preparation

README.md

@@ -41,15 +41,16 @@ The scalar FP32 decoder returns a global token ID.
 
 M3 MacBook Air (16 GB), macOS 27 build 26A428. Medians after warmup;
 frontend, encoder and decoding are included, while preparation, file I/O and
-chunk planning are excluded. The long JFK clip is 1,095.320125 seconds (18:15).
+chunk planning are excluded. The audio is JFK's **“We choose to go to the Moon”**
+speech: a **20-second excerpt** and the **18-minute 15-second recording**.
 The quality short-clip test intentionally uses the unmodified five-minute shape.
 
 | Bundle | Audio duration | Transcription | Audio / elapsed |
 | --- | ---: | ---: | ---: |
-| fast | 20.000 s | 0.150 s | 133.4× |
-| fast | 1095.320 s | 1.876 s | 584.0× |
-| quality | 20.000 s | 1.336 s | 15.0× |
-| quality | 1095.320 s | 7.044 s | 155.5× |
+| fast | 20 seconds | 0.150 s | 133.4× |
+| fast | 18 min 15 s | 1.876 s | 584.0× |
+| quality | 20 seconds | 1.336 s | 15.0× |
+| quality | 18 min 15 s | 7.044 s | 155.5× |
 
 Whisper-normalized WER against the supplied long-recording reference:
 
@@ -64,8 +65,22 @@ The reported full-recording runs retain all audio, with 74 fast chunks or four
 quality chunks. These are distinct contexts and execution policies, not a
 controlled kernel-only comparison.
 
-Source `.aimodel` files specialize on the device. First-use preparation can
-take several minutes; it is not included in the table. AoT compiled artifacts
+## Device specialization and loading
+
+Source `.aimodel` files specialize on the device. On the same Mac, preparation
+of these published source assets took:
+
+| Bundle | First observed preparation with specialization | Subsequent cached preparation |
+| --- | ---: | ---: |
+| fast | 33 s | 0.060 s |
+| quality | 10 min 7 s | 0.037 s |
+
+Preparation includes model loading and compilation requested by Core AI, and
+excludes warmup and transcription. These are observed cache histories, not
+guaranteed fresh-install times: the underlying ANE cache state is not fully
+observable, and OS/device/application changes can require specialization again.
+
+AoT compiled artifacts
 are omitted because no significant load-time benefit has been demonstrated.
 Authoring debug locations were removed while preserving graph signatures and
 operation counts. `SHA256.json` lists distributed payload checksums.

SHA256.json

@@ -11,8 +11,8 @@
   },
   {
     "path": "README.md",
-    "bytes": 3531,
-    "sha256": "d328e50a52bbbac7430e9cba2fd63c257a24fdf9c53af559e387b4c8cdf386c7"
+    "bytes": 4129,
+    "sha256": "e49a5b9f1ea0fa4d4fa8a369f6f9b732da44aae3733b55ea19b696fc384911dc"
   },
   {
     "path": "fast/LICENSE",