Add verified Raspberry Pi 5 edge performance measurements

Record actual 16 GB Pi 5 latency, RAM and correctness with frozen evaluation inputs, raw per-call measurements, runtime hashes and reproducible settings. Preserve weights and existing training data.

README.md

@@ -188,9 +188,9 @@ use a fresh independent evaluation set for new performance claims.
 
 ## Raspberry Pi evaluation snapshot — 2026-09-26
 
-Measurement/data revision: `edge-rpi5-20260926`. The original training/source files are unchanged, pinned at `8479d12161767c5aabd7dfc097e6f6c41790201a`. No training or fine-tuning was performed for this update.
+Measurement/data revision: `edge-rpi5-20260926.1`. The original training/source files are unchanged, pinned at `8479d12161767c5aabd7dfc097e6f6c41790201a`. No training or fine-tuning was performed for this update.
 
-The [`edge_benchmarks/rpi5-2026-09-26`](edge_benchmarks/rpi5-2026-09-26) directory adds **284 quality cases**, **10 latency inputs**, and all quality predictions from the corresponding published model on a Raspberry Pi 5 (16 GB, aarch64). The complete model card [reports the measurements](https://huggingface.co/sraivante/superfast-tiny-home-robotics-json-1m-v1/tree/edge-rpi5-20260926).
+The [`edge_benchmarks/rpi5-2026-09-26`](edge_benchmarks/rpi5-2026-09-26) directory adds **284 quality cases**, **10 latency inputs**, and all quality predictions from the corresponding published model on a Raspberry Pi 5 (16 GB, aarch64). The complete model card [reports the measurements](https://huggingface.co/sraivante/superfast-tiny-home-robotics-json-1m-v1/tree/edge-rpi5-20260926.1).
 
 Evaluation role: **existing frozen stress/regression suite, constrained decoding only**. Previously inspected synthetic suite; not a new blind holdout. Original per-case source-overlap metadata retained. The Pi result was **72/284**, with the per-category breakdown in the linked model report. This data does not establish fresh blind or real-world accuracy.
 

edge_benchmarks/rpi5-2026-09-26/provenance.json

@@ -1,5 +1,5 @@
 {
-  "measurement_revision": "edge-rpi5-20260926",
+  "measurement_revision": "edge-rpi5-20260926.1",
   "model": {
     "repo_id": "sraivante/superfast-tiny-home-robotics-json-1m-v1",
     "revision": "cf325b380e8ed560a729ec2de09f878a8a917259",

release_manifest.json

@@ -66,8 +66,8 @@
       "sha256": "3dce4a2870f76ab3b86cc84067bb017d32e8b726892fc87a8ca4110cac79219b"
     },
     "README.md": {
-      "bytes": 10531,
-      "sha256": "bfb74961ada4f61809e8ccc7adb80350ce68ecf4efd152735046a7f0d4255086"
+      "bytes": 10535,
+      "sha256": "8d85f545251cdcb1b0bae5fc31cd3c6b408a003951033a3068d9a237228db580"
     },
     "artifacts/home_commands_training_bundle.zip": {
       "bytes": 19263019,
@@ -98,8 +98,8 @@
       "sha256": "ce3493fe970f532956f1cbb487dceb143a0ab6dd029b25763c9bc27a9402ecca"
     },
     "edge_benchmarks/rpi5-2026-09-26/provenance.json": {
-      "bytes": 9358,
-      "sha256": "f5a12e015ad46ccd193afe2cd9aa7e2ed757384e07367578256fd328ea0e874c"
+      "bytes": 9360,
+      "sha256": "eea82a739be147d4e90fca76d92da40d3f8fc50d7fe11606f49be815f54ad36b"
     },
     "edge_benchmarks/rpi5-2026-09-26/quality_cases.jsonl": {
       "bytes": 56148,
@@ -169,7 +169,7 @@
   "documentation_patch": "Repository renamed at the publisher's request; model title, usage and linked dataset references updated. Weights, runtime code and all training/evaluation data are unchanged.",
   "previous_release_commit": "1fded12c78f5c9bd33775dd70c673d44753244ad",
   "edge_measurement_update": {
-    "revision": "edge-rpi5-20260926",
+    "revision": "edge-rpi5-20260926.1",
     "date": "2026-09-26",
     "weights_and_training_data_changed": false,
     "files": "edge_benchmarks/rpi5-2026-09-26"