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.
centaur
Characteristic dimensions (auto, from its own voice profile): ▲ S_MONO ▲ S_STRY ▲ S_NARR ▲ ESTH ▼ S_RANT ▼ EXPL ▼ DARC ▼ S_CONV · emo: Concentration Interest Emotional_Numbness
Training set = top-20% by reward = z(genu)+z(blend)+z(quality)−z(WER)+0.7·z(profile-match). Kept 127/636.
Best 5 of the kept top-20%
R 6.17 · genu 0.21 · blend 0.75 · qual 0.50 · WER 0.00
R 5.23 · genu 0.06 · blend 0.48 · qual 0.73 · WER 0.00
R 5.22 · genu 0.14 · blend 0.48 · qual 0.65 · WER 0.00
R 5.00 · genu 0.15 · blend 0.58 · qual 0.57 · WER 0.00
R 4.98 · genu 0.18 · blend 0.34 · qual 0.63 · WER 0.00
Weakest 5 still inside the top-20% (the selection boundary)
R 1.47 · genu 0.11 · blend 0.58 · qual 0.47 · WER 0.00
R 1.46 · genu 0.06 · blend 0.09 · qual 0.67 · WER 0.00
R 1.45 · genu 0.01 · blend 0.18 · qual 0.69 · WER 0.00
R 1.42 · genu 0.08 · blend 0.07 · qual 0.69 · WER 0.00
R 1.39 · genu 0.10 · blend 0.17 · qual 0.53 · 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.”