Add VibeVoice 7B result for the 4-person podcast spike

Same script/seed/voices as the 1.5B test, for direct comparison. Notably more content-faithful (no unscripted insertions), though one line drifted from a statement to a question. See vibevoice-spike/README.md. Co-Authored-By: Claude Sonnet 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01SN15Vxn4svKuhYxFsVtMJf
2026-09-30 22:04:13 +00:00
1 parent d1bd1ec955
commit f038ba609d
@@ -36,3 +36,17 @@ encouraging, not conclusive.
 Steves call after listening: pursue 7B (weights already downloaded, same
 venv/script, swap `--model_path`) or not worth it given the limitations
 the prior research already flagged.
+
+## 7B result (2026-09-30, same script/seed/voices)
+
+`trailer_promo_4speaker_vibevoice-7b.mp3` - VibeVoice-7B, same seed (42),
+same script, same reference voices. 34.8s of audio generated in 22.2s
+(RTF 0.64x - still faster than real-time, just less headroom than 1.5B).
+
+Whisper-transcript check: cleaner than 1.5B - neither of 1.5Bs unscripted
+insertions ("I am here!", "For me,") appeared, and "He will." came through
+exactly right this time (1.5B had drifted it to "You will."). One real
+deviation: Marges defensive line "I get one fact wrong per episode."
+came out as a question, "Can I get one fact wrong per episode?" - changes
+the delivery/meaning, not just wording. Still no audible disclaimer and
+nothing that reads as injected background music.