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.
abyssal-tyrant
Characteristic dimensions (auto, from its own voice profile): ▲ S_MONO ▲ S_STRY ▲ GEND ▲ S_NARR ▼ EXPL ▼ S_RANT ▼ S_CONV ▼ DARC · 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 123/615.
Best 5 of the kept top-20%
R 6.75 · genu 0.45 · blend 0.25 · qual 0.43 · WER 0.00
R 6.20 · genu 0.42 · blend 0.29 · qual 0.41 · WER 0.06
R 4.51 · genu 0.13 · blend 0.52 · qual 0.56 · WER 0.00
R 4.19 · genu 0.17 · blend 0.26 · qual 0.62 · WER 0.00
R 4.19 · genu 0.19 · blend 0.31 · qual 0.64 · WER 0.00
Weakest 5 still inside the top-20% (the selection boundary)
R 1.21 · genu 0.17 · blend 0.21 · qual 0.44 · WER 0.03
R 1.20 · genu 0.02 · blend 0.00 · qual 0.73 · WER 0.00
R 1.20 · genu 0.14 · blend 0.18 · qual 0.54 · WER 0.03
R 1.19 · genu 0.20 · blend 0.48 · qual 0.40 · WER 0.53
R 1.18 · genu 0.08 · blend 0.08 · qual 0.54 · 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.”