perf(pine): whisper --beam-size 1 — ~45% faster STT, same transcripts

Image default is beam 5 on x86 (1 on ARM). Benchmarked on framework-pt
(i5-1135G7, base-int8, real speech): beam 1 transcribes identical text
~45% faster; extra CPU threads made it slower, so only the beam changes.
App revision 3.4.2 > image 3.4.1 so catalog-driven nodes roll the args
change; a 3.4.1-1 suffix would semver-compare LOWER and never ship.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

apps/pine-whisper/manifest.yml

@@ -1,7 +1,11 @@
 app:
   id: pine-whisper
   name: Pine Whisper (STT)
-  version: "3.4.1"
+  # App revision 3.4.2 = upstream wyoming-whisper 3.4.1 image + tuned args
+  # (--beam-size 1). Bumped past the image version so catalog-driven nodes
+  # pick up the args change; the pre-release form "3.4.1-1" would compare
+  # LOWER than 3.4.1 under semver and never roll out.
+  version: "3.4.2"
   description: Wyoming-protocol faster-whisper speech-to-text engine. Internal Pine voice-assistant stack member — turns speech captured by a PineVoice satellite into text for Home Assistant Assist.
   category: home
 
@@ -18,7 +22,11 @@ app:
     network_aliases: [pine-whisper]
     # The image entrypoint already binds tcp://0.0.0.0:10300; these args only
     # pick the model + language (mirrors the pine ha-stack.yml compose command).
-    custom_args: ["--model", "base-int8", "--language", "en"]
+    # --beam-size 1: the image default is 5 on x86 (1 on ARM). Benchmarked on
+    # framework-pt (i5-1135G7, base-int8): beam 1 transcribes the same text
+    # ~45% faster — the standard low-latency setting for short voice commands
+    # (HA's own whisper add-on defaults to 1).
+    custom_args: ["--model", "base-int8", "--language", "en", "--beam-size", "1"]
 
   dependencies:
     - storage: 2Gi