10-generation evolution of prompt + LoRA-strength + sampling to maximize Contentment. Fitness = mean over 8 candidates. Best evolved 0.129 vs prompt-without-LoRA 0.000 (LoRA lift +0.129) vs neutral 0.008; winning λ≈1.0.
Method & reproducibility — how the evolution ran
Base model:laion/moss-tts-local-transformer-4.55b-voice-acting-v2 · LoRA: the matching v3 emotion LoRA (TTS-AGI/moss-emotion-loras-v3, rank 32), applied via PEFT adapter scaling at strength λ. Algorithm: genetic search — population 16, 10 generations, elitism (top 6 carried over). Each generation: every genome is generated best-of-8 and its fitness = the mean target-emotion strength over those 8 candidates (EmoNet score for the emotion, minus 0.25·max(0, WER−0.6) as a light intelligibility guard). Offspring = 60% crossover of two elites + mutation, 40% mutation of one elite. Genome (what evolves): the GENERAL voice description, the SCRIPT delivery cue, the carrier sentence, the LoRA strength λ (seeded 0.75–1.5, mutated within 0.4–2.0), and the sampling params (audio temperature, top_p, top_k). Mutation operators: 35% regenerate the description (new intensity + manner adjectives), 20% swap the carrier sentence, 23% jitter λ by ±0.25/0.5, 12% jitter temperature ±0.1, 10% change top_k. LoRA on vs off: the two LoRA OFF cards are baselines generated with the adapter disabled (λ=0) — a plain neutral prompt, and the best steering prompt without the LoRA. Every LoRA ON hall-of-fame recipe below was generated with the LoRA active at the evolved λ shown on the card. Seed is derived from the emotion name (deterministic).
Baselines — LoRA disabled
NEUTRAL promptLoRA OFF (λ=0)
mean-emo 0.008 · best 0.04 · blend 0.21 · genu 0.05 · WER 0.47
STEER promptLoRA OFF (λ=0)
mean-emo 0.000 · best 0.00 · blend 0.27 · genu 0.09 · WER 0.25
GENERAL: A voice to the extreme expressing contentment, a relaxed, satisfied voice, at peace and content, easy and unhurried, pouring out.
SCRIPT: (calmly content and at ease) "The report is on the desk, right next to the printer."
Hall of fame — top 6 evolved recipes (all generated WITH the emotion LoRA active)
#1LoRA ON · λ=1.0 (100%)
mean-emo 0.129 · best 0.32 · blend 0.35 · genu 0.03 · WER 0.08
GENERAL: A voice utterly expressing contentment, a relaxed, satisfied voice, at peace and content, easy and unhurried, in every breath.
SCRIPT: (calmly content and at ease) "There are twelve chairs arranged around the conference table."
mean-emo 0.128 · best 0.37 · blend 0.31 · genu 0.07 · WER 0.07
GENERAL: A voice deeply expressing contentment, a relaxed, satisfied voice, at peace and content, easy and unhurried, pouring out.
SCRIPT: (calmly content and at ease) "There are twelve chairs arranged around the conference table."
mean-emo 0.129 · best 0.27 · blend 0.26 · genu 0.04 · WER 0.61
GENERAL: A voice unmistakably expressing contentment, a relaxed, satisfied voice, at peace and content, easy and unhurried, in every breath.
SCRIPT: (calmly content and at ease) "There are twelve chairs arranged around the conference table."
mean-emo 0.121 · best 0.36 · blend 0.26 · genu 0.07 · WER 0.33
GENERAL: A voice deeply expressing contentment, a relaxed, satisfied voice, at peace and content, easy and unhurried, raw and unfiltered.
SCRIPT: (calmly content and at ease) "There are twelve chairs arranged around the conference table."
mean-emo 0.111 · best 0.31 · blend 0.26 · genu 0.04 · WER 0.07
GENERAL: A voice utterly expressing contentment, a relaxed, satisfied voice, at peace and content, easy and unhurried, in every breath.
SCRIPT: (calmly content and at ease) "There are twelve chairs arranged around the conference table."
mean-emo 0.104 · best 0.31 · blend 0.30 · genu 0.09 · WER 0.17
GENERAL: A voice unmistakably expressing contentment, a relaxed, satisfied voice, at peace and content, easy and unhurried, in every breath.
SCRIPT: (calmly content and at ease) "There are twelve chairs arranged around the conference table."
