The merged parent used in these specialist comparisons is now released as Humaneness Voice Large 1.6 Alpha: 1.5 + final RealSpeech rank 4, merged once. Apply only the selected new bucket adapter. Do not reload either ancestor.
1.5 Original model · 1.6 merged model · Complete context manuals for humans and agents
Read 1.5 context manual · Read 1.6 + adapter context manual. Exact prompts, CFG, adapter scales and failed gates are retained. New model/documentation contributions: CC BY 4.0; existing upstream and gated specialist terms remain applicable.
The original grid used generic GENERAL/SCRIPT directions at CFG1. It did not systematically use target-specific manual winners. The new study freezes the actual rank-4 manual wording, preserves every literal word, seed, reference, adapter and frame budget, then independently varies CFG and adapter strength. The previous recordings and conclusions remain below.
Ranking:100% requested EARS emotion/style head, standardized using frozen training statistics; low-pole targets reverse the sign. Formal/casual contrasts average the desired head and negative opposite head. CE and genuineness are diagnostics. Higher target z means stronger desired delivery, not a human rating. More CFG is not automatically better.
Selection:20 original-text outcomes per configuration select wording, CFG and scale. Fifty additional unseen-text reference pairs test the frozen choices. A recommended adapter must improve against both the same-CFG no-adapter control and the best manual-only setting. Quality gates: per-take WER≤15%, mean WER≤10%, valid/terminated fraction≥90%, paired identity change≥−0.05 over50pairs. Failed outcomes stay visible. Language-cluster bootstrap intervals use only five fixed reference voices; findings across72targets remain exploratory.
Five languages: EN/DE/FR/NL/ES; original seeds1701/1702; reference187frames/14.96seconds. CFG1/1.25/1.5/2/2.5/3/4/5 × bucket strength0/0.5/0.75/1. CFG4/5 are stress conditions, not presumed improvements. No-reference use of these bucket recommendations has not been validated.
Taxonomy:S_MONO means Monologue, S_TECH means Teacher/Didactic, VOLT means Volatility. The two corrected style recipes and17dimension/contrast recipes are new searches, not historical manual winners.53other targets reuse directly applicable manual recipes.
Best tested agent route: Use manual-only configuration; no clear qualifying incremental adapter gain. Bucket scale 0.0, CFG 3.0.
| Held-out condition | n | Target z↑ | Raw head | CE↑ | Genuine↑ | Voice↑ | WER↓ | Valid | QA |
|---|---|---|---|---|---|---|---|---|---|
| standard | 50 | 0.225 | 0.402 | 6.257 | 4.314 | 0.555 | 0.008 | 1.000 | pass |
| manual | 50 | 2.787 | 1.817 | 6.138 | 4.118 | 0.525 | 0.007 | 1.000 | pass |
| adapter | 50 | 2.804 | 1.826 | 6.151 | 4.170 | 0.516 | 0.006 | 1.000 | pass |
| matched_no_adapter | 50 | 2.787 | 1.817 | 6.138 | 4.118 | 0.525 | 0.007 | 1.000 | pass |
Incremental adapter gain versus same-CFG control: 0.017 z;95% cluster interval [-0.33966795072896655, 0.2641015536117598]. Versus best manual-only: 0.017 z. Manual-only versus standard: 2.562 z.
| Scale | CFG | Target z | CE | Genuine | WER | Provisional QA |
|---|---|---|---|---|---|---|
| 0.0 | 1.0 | 0.379 | 6.175 | 4.311 | 0.000 | True |
| 0.5 | 1.0 | 0.284 | 6.198 | 4.317 | 0.005 | True |
| 0.75 | 1.0 | 0.377 | 6.207 | 4.346 | 0.005 | True |
| 1.0 | 1.0 | 0.248 | 6.180 | 4.329 | 0.005 | True |
| 0.0 | 1.25 | 0.579 | 6.202 | 4.345 | 0.036 | True |
| 0.5 | 1.25 | 0.526 | 6.168 | 4.295 | 0.000 | True |
| 0.75 | 1.25 | 0.754 | 6.208 | 4.338 | 0.000 | True |
| 1.0 | 1.25 | 0.516 | 6.203 | 4.400 | 0.003 | True |
| 0.0 | 1.5 | 0.738 | 6.181 | 4.358 | 0.000 | True |
| 0.5 | 1.5 | 0.879 | 6.199 | 4.333 | 0.008 | True |
| 0.75 | 1.5 | 1.056 | 6.176 | 4.350 | 0.005 | True |
| 1.0 | 1.5 | 0.820 | 6.218 | 4.362 | 0.000 | True |
| 0.0 | 2.0 | 1.821 | 6.161 | 4.266 | 0.003 | True |
| 0.5 | 2.0 | 1.567 | 6.159 | 4.264 | 0.003 | True |
| 0.75 | 2.0 | 1.221 | 6.176 | 4.222 | 0.003 | True |
| 1.0 | 2.0 | 1.587 | 6.161 | 4.284 | 0.003 | True |
| 0.0 | 2.5 | 2.156 | 6.151 | 4.271 | 0.003 | True |
| 0.5 | 2.5 | 1.936 | 6.107 | 4.229 | 0.005 | True |
| 0.75 | 2.5 | 2.357 | 6.127 | 4.200 | 0.003 | False |
| 1.0 | 2.5 | 2.121 | 6.100 | 4.239 | 0.000 | True |
| 0.0 | 3.0 | 2.969 | 6.113 | 4.184 | 0.008 | True |
| 0.5 | 3.0 | 3.021 | 6.058 | 4.184 | 0.003 | True |
| 0.75 | 3.0 | 2.599 | 6.092 | 4.150 | 0.000 | True |
| 1.0 | 3.0 | 2.893 | 6.041 | 4.120 | 0.003 | True |
| 0.0 | 4.0 | 3.959 | 6.010 | 4.037 | 0.000 | False |
| 0.5 | 4.0 | 4.044 | 6.001 | 4.023 | 0.000 | False |
| 0.75 | 4.0 | 3.830 | 6.000 | 4.051 | 0.006 | False |
| 1.0 | 4.0 | 3.671 | 5.987 | 4.098 | 0.005 | False |
| 0.0 | 5.0 | 4.638 | 5.897 | 3.947 | 0.025 | False |
| 0.5 | 5.0 | 4.845 | 5.863 | 3.898 | 0.000 | False |
| 0.75 | 5.0 | 4.655 | 5.957 | 3.918 | 0.006 | False |
| 1.0 | 5.0 | 4.700 | 5.837 | 3.942 | 0.021 | False |
direct_manual_recipe
{
"id": "manual-bedaa97ab879",
"general": "Target resonance shift; the surprise should sound as if the vocal tract has momentarily widened or constricted drastically, changing the perceived timbre.",
"inline": [
"Deliver the opening with a tightly focused, forward placement in the mouth, low in resonance.",
"Immediately shift the resonance upward and backward, creating a sudden, bright, almost metallic quality of surprise.",
"Let the final articulation carry that heightened, bright resonance, sustaining the unusual timbre until the very end."
],
"cfg": 3,
"recipe": "P05",
"strategy": "local-Gemma-mutation",
"timing_syntax": "canonical",
"canonical_labels": true,
"neutral_event": "keep",
"parent": "cfg-3",
"origin": "exact_frozen_manual_recipe"
}Exact machine-readable comparison · Agent tool recommendations
Intense Astonishment Surprise. Base: 1.5 Alpha + merged real-speech rank4; new adapter: rank32. Every row is a matched seed/text/reference test. Scale zero disables the new adapter; all other scales are applied absolutely, never compounded.
Adapter training data: {"es": 9, "de": 16, "en": 24, "nl": 9, "fr": 8}. All available accepted languages were pooled, globally shuffled, for three epochs. Peak LR 2e-05 at the end of the first-epoch warmup, then cosine decay. All original optimizer/RNG checkpoints remain saved.
Taxonomy: S_MONO is monologue, not monotone; VOLT is volatility, not volume. No claim is made that a classifier score alone proves the audible performance.
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 20 | 0.518 ± 0.211 | 6.207 ± 0.128 | 4.367 ± 0.205 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 20 | 0.547 ± 0.234 | 6.226 ± 0.113 | 4.354 ± 0.226 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 20 | 0.612 ± 0.250 | 6.209 ± 0.108 | 4.302 ± 0.241 | 0.29% ± 1.32 |
| Rank32 × 1.0 | 20 | 0.561 ± 0.210 | 6.168 ± 0.209 | 4.320 ± 0.235 | 0.00% ± 0.00 |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.675 ± 0.325 | 6.299 ± 0.121 | 4.206 ± 0.180 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.660 ± 0.320 | 6.269 ± 0.172 | 4.187 ± 0.267 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.711 ± 0.328 | 6.247 ± 0.066 | 4.081 ± 0.285 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.647 ± 0.348 | 6.318 ± 0.073 | 4.179 ± 0.235 | 0.00% ± 0.00 |
| Prompt / seed | Merged base · no new LoRA | Rank32 × 0.5 | Rank32 × 0.75 | Rank32 × 1.0 |
|---|---|---|---|---|
| Seed 1701 The door is still open. I put your coat on the chair. Tell me when you are ready. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready. | Target 0.449 · CE 6.21 · Genuine 4.08 · WER 0.0% | Target 0.338 · CE 6.34 · Genuine 4.07 · WER 0.0% | Target 0.428 · CE 6.24 · Genuine 3.90 · WER 0.0% | Target 0.357 · CE 6.28 · Genuine 4.06 · WER 0.0% |
| Seed 1702 The door is still open. I put your coat on the chair. Tell me when you are ready. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready. | Target 0.345 · CE 6.19 · Genuine 4.06 · WER 0.0% | Target 0.436 · CE 6.02 · Genuine 3.87 · WER 0.0% | Target 0.426 · CE 6.16 · Genuine 3.79 · WER 0.0% | Target 0.335 · CE 6.24 · Genuine 3.90 · WER 0.0% |
| Seed 1701 I found the little box. It was beside the old window. We can look inside together. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together. | Target 0.990 · CE 6.36 · Genuine 4.24 · WER 0.0% | Target 0.983 · CE 6.29 · Genuine 4.41 · WER 0.0% | Target 0.995 · CE 6.30 · Genuine 4.40 · WER 0.0% | Target 0.966 · CE 6.39 · Genuine 4.37 · WER 0.0% |
| Seed 1702 I found the little box. It was beside the old window. We can look inside together. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together. | Target 0.917 · CE 6.44 · Genuine 4.44 · WER 0.0% | Target 0.881 · CE 6.42 · Genuine 4.41 · WER 0.0% | Target 0.996 · CE 6.29 · Genuine 4.23 · WER 0.0% | Target 0.930 · CE 6.37 · Genuine 4.38 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.467 ± 0.063 | 6.309 ± 0.092 | 4.357 ± 0.089 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.390 ± 0.056 | 6.336 ± 0.061 | 4.372 ± 0.165 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.542 ± 0.291 | 6.291 ± 0.120 | 4.452 ± 0.235 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.455 ± 0.109 | 6.316 ± 0.185 | 4.512 ± 0.243 | 0.00% ± 0.00 |
| Prompt / seed | Merged base · no new LoRA | Rank32 × 0.5 | Rank32 × 0.75 | Rank32 × 1.0 |
|---|---|---|---|---|
| Seed 1701 Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist. | Target 0.548 · CE 6.24 · Genuine 4.31 · WER 0.0% | Target 0.349 · CE 6.37 · Genuine 4.55 · WER 0.0% | Target 0.357 · CE 6.21 · Genuine 4.57 · WER 0.0% | Target 0.389 · CE 6.10 · Genuine 4.78 · WER 0.0% |
| Seed 1702 Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist. | Target 0.426 · CE 6.22 · Genuine 4.48 · WER 0.0% | Target 0.342 · CE 6.25 · Genuine 4.47 · WER 0.0% | Target 0.353 · CE 6.17 · Genuine 4.71 · WER 0.0% | Target 0.364 · CE 6.24 · Genuine 4.20 · WER 0.0% |
| Seed 1701 Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen. | Target 0.409 · CE 6.38 · Genuine 4.36 · WER 0.0% | Target 0.411 · CE 6.36 · Genuine 4.29 · WER 0.0% | Target 0.969 · CE 6.42 · Genuine 4.36 · WER 0.0% | Target 0.607 · CE 6.40 · Genuine 4.58 · WER 0.0% |
| Seed 1702 Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen. | Target 0.484 · CE 6.40 · Genuine 4.28 · WER 0.0% | Target 0.460 · CE 6.36 · Genuine 4.18 · WER 0.0% | Target 0.489 · CE 6.36 · Genuine 4.17 · WER 0.0% | Target 0.460 · CE 6.52 · Genuine 4.48 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.357 ± 0.082 | 6.128 ± 0.146 | 4.583 ± 0.102 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.332 ± 0.020 | 6.187 ± 0.034 | 4.589 ± 0.083 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.413 ± 0.173 | 6.185 ± 0.106 | 4.376 ± 0.155 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.500 ± 0.207 | 6.023 ± 0.338 | 4.385 ± 0.137 | 0.00% ± 0.00 |
| Prompt / seed | Merged base · no new LoRA | Rank32 × 0.5 | Rank32 × 0.75 | Rank32 × 1.0 |
|---|---|---|---|---|
| Seed 1701 La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt. | Target 0.324 · CE 6.24 · Genuine 4.63 · WER 0.0% | Target 0.309 · CE 6.17 · Genuine 4.48 · WER 0.0% | Target 0.332 · CE 6.11 · Genuine 4.14 · WER 0.0% | Target 0.416 · CE 6.00 · Genuine 4.38 · WER 0.0% |
| Seed 1702 La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt. | Target 0.307 · CE 6.26 · Genuine 4.43 · WER 0.0% | Target 0.326 · CE 6.15 · Genuine 4.68 · WER 0.0% | Target 0.352 · CE 6.08 · Genuine 4.44 · WER 0.0% | Target 0.459 · CE 6.25 · Genuine 4.44 · WER 0.0% |
| Seed 1701 J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient. | Target 0.479 · CE 6.05 · Genuine 4.61 · WER 0.0% | Target 0.357 · CE 6.23 · Genuine 4.58 · WER 0.0% | Target 0.671 · CE 6.28 · Genuine 4.46 · WER 0.0% | Target 0.326 · CE 6.28 · Genuine 4.52 · WER 0.0% |
| Seed 1702 J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient. | Target 0.317 · CE 5.96 · Genuine 4.66 · WER 0.0% | Target 0.335 · CE 6.20 · Genuine 4.61 · WER 0.0% | Target 0.296 · CE 6.27 · Genuine 4.46 · WER 0.0% | Target 0.798 · CE 5.55 · Genuine 4.20 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.558 ± 0.229 | 6.132 ± 0.093 | 4.230 ± 0.079 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.685 ± 0.188 | 6.179 ± 0.016 | 4.261 ± 0.077 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.731 ± 0.145 | 6.170 ± 0.121 | 4.322 ± 0.205 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.639 ± 0.159 | 6.106 ± 0.031 | 4.263 ± 0.156 | 0.00% ± 0.00 |
| Prompt / seed | Merged base · no new LoRA | Rank32 × 0.5 | Rank32 × 0.75 | Rank32 × 1.0 |
|---|---|---|---|---|
| Seed 1701 De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent. | Target 0.834 · CE 6.01 · Genuine 4.12 · WER 0.0% | Target 0.948 · CE 6.20 · Genuine 4.31 · WER 0.0% | Target 0.729 · CE 6.01 · Genuine 4.09 · WER 0.0% | Target 0.758 · CE 6.15 · Genuine 4.17 · WER 0.0% |
| Seed 1702 De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent. | Target 0.658 · CE 6.23 · Genuine 4.31 · WER 0.0% | Target 0.568 · CE 6.19 · Genuine 4.32 · WER 0.0% | Target 0.534 · CE 6.31 · Genuine 4.58 · WER 0.0% | Target 0.676 · CE 6.08 · Genuine 4.43 · WER 0.0% |
| Seed 1701 Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken. | Target 0.378 · CE 6.16 · Genuine 4.25 · WER 0.0% | Target 0.532 · CE 6.17 · Genuine 4.15 · WER 0.0% | Target 0.780 · CE 6.20 · Genuine 4.36 · WER 0.0% | Target 0.717 · CE 6.10 · Genuine 4.35 · WER 0.0% |
| Seed 1702 Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken. | Target 0.363 · CE 6.13 · Genuine 4.24 · WER 0.0% | Target 0.694 · CE 6.17 · Genuine 4.27 · WER 0.0% | Target 0.879 · CE 6.16 · Genuine 4.26 · WER 0.0% | Target 0.406 · CE 6.09 · Genuine 4.09 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.531 ± 0.199 | 6.168 ± 0.086 | 4.459 ± 0.273 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.666 ± 0.224 | 6.160 ± 0.129 | 4.361 ± 0.301 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.665 ± 0.242 | 6.154 ± 0.107 | 4.278 ± 0.247 | 1.47% ± 2.94 |
| Rank32 × 1.0 | 4 | 0.562 ± 0.202 | 6.078 ± 0.136 | 4.259 ± 0.321 | 0.00% ± 0.00 |
| Prompt / seed | Merged base · no new LoRA | Rank32 × 0.5 | Rank32 × 0.75 | Rank32 × 1.0 |
|---|---|---|---|---|
| Seed 1701 La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo. | Target 0.376 · CE 6.07 · Genuine 4.19 · WER 0.0% | Target 0.334 · CE 6.17 · Genuine 4.12 · WER 0.0% | Target 0.430 · CE 6.28 · Genuine 4.42 · WER 0.0% | Target 0.625 · CE 6.18 · Genuine 4.24 · WER 0.0% |
| Seed 1702 La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo. | Target 0.371 · CE 6.13 · Genuine 4.29 · WER 0.0% | Target 0.758 · CE 5.97 · Genuine 4.08 · WER 0.0% | Target 0.511 · CE 6.02 · Genuine 3.93 · WER 0.0% | Target 0.467 · CE 5.90 · Genuine 3.84 · WER 0.0% |
| Seed 1701 He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro. | Target 0.785 · CE 6.21 · Genuine 4.56 · WER 0.0% | Target 0.752 · CE 6.27 · Genuine 4.58 · WER 0.0% | Target 0.963 · CE 6.19 · Genuine 4.48 · WER 0.0% | Target 0.811 · CE 6.05 · Genuine 4.33 · WER 0.0% |
| Seed 1702 He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro. Exact GENERAL / SCRIPT promptGENERAL: One adult speaker in a clean recording. Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja. (Intense Astonishment Surprise. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro. | Target 0.594 · CE 6.26 · Genuine 4.80 · WER 0.0% | Target 0.821 · CE 6.22 · Genuine 4.66 · WER 0.0% | Target 0.755 · CE 6.13 · Genuine 4.28 · WER 5.9% | Target 0.345 · CE 6.18 · Genuine 4.62 · WER 0.0% |