Profile-reward recipe: generate 640 (batched) → ASR-filter+trim → SIDON → keep the top 20% by a per-character quality+profile reward → warm-start FT 3 epochs. Below: the quality range of the selected training set, then example generations after each epoch.
skeleton
Characteristic dimensions (auto, from its own voice profile): ▲ S_STRY ▲ S_MONO ▲ S_NARR ▲ S_DRAM ▼ EXPL ▼ S_NEWS ▼ S_TECH ▼ S_CONV · emo: Concentration Emotional_Numbness Interest
Training set = top-20% by reward = z(genu)+z(blend)+z(quality)−z(WER)+0.7·z(profile-match). Kept 126/632.
Best 5 of the kept top-20%
R 7.84 · genu 0.41 · blend 0.64 · qual 0.36 · WER 0.03
R 6.18 · genu 0.19 · blend 0.41 · qual 0.66 · WER 0.00
R 4.35 · genu 0.17 · blend 0.41 · qual 0.53 · WER 0.00
R 4.12 · genu 0.13 · blend 0.63 · qual 0.53 · WER 0.00
R 4.11 · genu 0.16 · blend 0.57 · qual 0.53 · WER 0.00
Weakest 5 still inside the top-20% (the selection boundary)
R 1.37 · genu 0.22 · blend 0.11 · qual 0.44 · WER 0.00
R 1.35 · genu 0.12 · blend 0.17 · qual 0.52 · WER 0.00
R 1.32 · genu 0.06 · blend 0.01 · qual 0.67 · WER 0.00
R 1.31 · genu 0.05 · blend 0.11 · qual 0.60 · WER 0.00
R 1.31 · genu 0.00 · blend 0.09 · qual 0.69 · WER 0.00
Fantasy line — refined ep1 / ep2 / ep3
“The old road winds past the ruined tower where the last dragon knights once kept their watch.”
Sci-fi line — refined ep1 / ep2 / ep3
“The station drifted in low orbit while the reactor cooled and the scanners swept the empty dark.”
Neutral line — refined ep1 / ep2 / ep3
“The meeting is scheduled for three o'clock and the report is on the desk by the printer.”