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.
gravelly-mountain-elder
Characteristic dimensions (auto, from its own voice profile): ▲ S_STRY ▲ S_MONO ▲ ESTH ▲ STRU ▼ S_RANT ▼ EXPL ▼ S_TECH ▼ DARC · emo: Concentration Interest Awe
Training set = top-20% by reward = z(genu)+z(blend)+z(quality)−z(WER)+0.7·z(profile-match). Kept 127/637.
Best 5 of the kept top-20%
R 7.80 · genu 0.54 · blend 0.23 · qual 0.41 · WER 0.00
R 6.43 · genu 0.27 · blend 0.25 · qual 0.66 · WER 0.00
R 5.70 · genu 0.20 · blend 0.34 · qual 0.66 · WER 0.00
R 5.68 · genu 0.27 · blend 0.56 · qual 0.37 · WER 0.00
R 5.00 · genu 0.20 · blend 0.37 · qual 0.60 · WER 0.03
Weakest 5 still inside the top-20% (the selection boundary)
R 1.40 · genu 0.00 · blend 0.01 · qual 0.78 · WER 0.00
R 1.39 · genu 0.00 · blend 0.14 · qual 0.64 · WER 0.00
R 1.39 · genu 0.14 · blend 0.00 · qual 0.60 · WER 0.00
R 1.39 · genu 0.03 · blend 0.04 · qual 0.69 · WER 0.00
R 1.39 · genu 0.06 · blend 0.16 · qual 0.56 · 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.”