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 4.0.
| Held-out condition | n | Target z↑ | Raw head | CE↑ | Genuine↑ | Voice↑ | WER↓ | Valid | QA |
|---|---|---|---|---|---|---|---|---|---|
| standard | 50 | -0.480 | 0.000 | 6.263 | 4.362 | 0.578 | 0.006 | 1.000 | pass |
| manual | 50 | -0.460 | 0.010 | 5.977 | 4.542 | 0.588 | 0.024 | 0.980 | pass |
| adapter | 50 | -0.462 | 0.009 | 5.956 | 4.520 | 0.594 | 0.041 | 0.920 | pass |
| matched_no_adapter | 50 | -0.460 | 0.010 | 5.977 | 4.542 | 0.588 | 0.024 | 0.980 | pass |
Incremental adapter gain versus same-CFG control: -0.002 z;95% cluster interval [-0.008828122707002494, 0.005742194256246636]. Versus best manual-only: -0.002 z. Manual-only versus standard: 0.021 z.
| Scale | CFG | Target z | CE | Genuine | WER | Provisional QA |
|---|---|---|---|---|---|---|
| 0.0 | 1.0 | -0.479 | 6.192 | 4.436 | 0.000 | True |
| 0.5 | 1.0 | -0.475 | 6.184 | 4.457 | 0.000 | True |
| 0.75 | 1.0 | -0.467 | 6.152 | 4.392 | 0.003 | True |
| 1.0 | 1.0 | -0.474 | 6.138 | 4.370 | 0.000 | True |
| 0.0 | 1.25 | -0.479 | 6.148 | 4.359 | 0.000 | True |
| 0.5 | 1.25 | -0.469 | 6.114 | 4.385 | 0.005 | True |
| 0.75 | 1.25 | -0.471 | 6.145 | 4.371 | 0.000 | True |
| 1.0 | 1.25 | -0.458 | 6.143 | 4.453 | 0.000 | True |
| 0.0 | 1.5 | -0.465 | 6.127 | 4.441 | 0.000 | True |
| 0.5 | 1.5 | -0.463 | 6.113 | 4.417 | 0.003 | True |
| 0.75 | 1.5 | -0.469 | 6.132 | 4.474 | 0.003 | True |
| 1.0 | 1.5 | -0.469 | 6.110 | 4.447 | 0.003 | True |
| 0.0 | 2.0 | -0.463 | 6.110 | 4.484 | 0.006 | True |
| 0.5 | 2.0 | -0.458 | 6.053 | 4.441 | 0.000 | True |
| 0.75 | 2.0 | -0.457 | 5.990 | 4.513 | 0.014 | True |
| 1.0 | 2.0 | -0.464 | 6.062 | 4.441 | 0.006 | True |
| 0.0 | 2.5 | -0.462 | 6.045 | 4.480 | 0.003 | True |
| 0.5 | 2.5 | -0.460 | 6.062 | 4.529 | 0.000 | True |
| 0.75 | 2.5 | -0.456 | 5.983 | 4.493 | 0.012 | True |
| 1.0 | 2.5 | -0.455 | 5.990 | 4.460 | 0.000 | True |
| 0.0 | 3.0 | -0.457 | 5.935 | 4.460 | 0.003 | True |
| 0.5 | 3.0 | -0.453 | 5.938 | 4.513 | 0.000 | True |
| 0.75 | 3.0 | -0.456 | 5.916 | 4.444 | 0.000 | True |
| 1.0 | 3.0 | -0.451 | 5.919 | 4.502 | 0.006 | True |
| 0.0 | 4.0 | -0.440 | 5.851 | 4.479 | 0.016 | True |
| 0.5 | 4.0 | -0.444 | 5.773 | 4.430 | 0.006 | True |
| 0.75 | 4.0 | -0.452 | 5.830 | 4.469 | 0.012 | True |
| 1.0 | 4.0 | -0.442 | 5.781 | 4.483 | 0.016 | True |
| 0.0 | 5.0 | -0.444 | 5.692 | 4.440 | 0.009 | True |
| 0.5 | 5.0 | -0.447 | 5.689 | 4.413 | 0.015 | True |
| 0.75 | 5.0 | -0.450 | 5.618 | 4.377 | 0.026 | True |
| 1.0 | 5.0 | -0.443 | 5.700 | 4.435 | 0.018 | True |
direct_manual_recipe
{
"id": "manual-c1306bd27b08",
"general": "Adopt a heavy, resonant vocal quality that sounds deliberately slow and weighty, using pauses as weapons to emphasize underlying suspicion and control.",
"inline": [
"Begin with a deep, deliberate breath that forces a slow, measured enunciation of the first phrase.",
"Insert a significant, pregnant pause mid-thought, letting the silence itself communicate the weight of the perceived slight.",
"Conclude the statement with a low, drawn-out vowel sound that trails off, suggesting unresolved resentment."
],
"cfg": 3,
"recipe": "P02",
"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 Jealousy and Envy. 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: {"fr": 6, "es": 7, "en": 5, "de": 11, "nl": 1}. 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 | 0.318 ± 0.010 | 6.183 ± 0.146 | 4.339 ± 0.266 | 0.31% ± 1.40 |
| Rank32 × 0.5 | 20 | 0.315 ± 0.007 | 6.215 ± 0.107 | 4.279 ± 0.225 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 20 | 0.316 ± 0.010 | 6.212 ± 0.117 | 4.385 ± 0.208 | 1.67% ± 7.45 |
| Rank32 × 1.0 | 20 | 0.314 ± 0.008 | 6.224 ± 0.105 | 4.375 ± 0.164 | 0.00% ± 0.00 |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.313 ± 0.006 | 6.336 ± 0.082 | 4.101 ± 0.208 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.309 ± 0.009 | 6.269 ± 0.064 | 4.118 ± 0.364 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.315 ± 0.008 | 6.336 ± 0.071 | 4.272 ± 0.185 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.316 ± 0.010 | 6.269 ± 0.062 | 4.266 ± 0.191 | 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.314 · CE 6.32 · Genuine 3.97 · WER 0.0% | Target 0.318 · CE 6.26 · Genuine 4.10 · WER 0.0% | Target 0.319 · CE 6.29 · Genuine 4.02 · WER 0.0% | Target 0.314 · CE 6.18 · Genuine 4.01 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.318 · CE 6.23 · Genuine 3.88 · WER 0.0% | Target 0.314 · CE 6.18 · Genuine 3.61 · WER 0.0% | Target 0.324 · CE 6.34 · Genuine 4.33 · WER 0.0% | Target 0.331 · CE 6.27 · Genuine 4.24 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.313 · CE 6.41 · Genuine 4.28 · WER 0.0% | Target 0.306 · CE 6.32 · Genuine 4.39 · WER 0.0% | Target 0.313 · CE 6.44 · Genuine 4.46 · WER 0.0% | Target 0.311 · CE 6.32 · Genuine 4.43 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.304 · CE 6.39 · Genuine 4.28 · WER 0.0% | Target 0.299 · CE 6.32 · Genuine 4.37 · WER 0.0% | Target 0.304 · CE 6.28 · Genuine 4.28 · WER 0.0% | Target 0.310 · CE 6.31 · Genuine 4.39 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.327 ± 0.012 | 6.297 ± 0.107 | 4.371 ± 0.184 | 1.56% ± 3.12 |
| Rank32 × 0.5 | 4 | 0.322 ± 0.004 | 6.275 ± 0.084 | 4.277 ± 0.170 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.329 ± 0.011 | 6.326 ± 0.079 | 4.435 ± 0.163 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.319 ± 0.012 | 6.350 ± 0.090 | 4.401 ± 0.173 | 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.327 · CE 6.32 · Genuine 4.55 · WER 0.0% | Target 0.326 · CE 6.17 · Genuine 4.52 · WER 0.0% | Target 0.341 · CE 6.27 · Genuine 4.56 · WER 0.0% | Target 0.333 · CE 6.28 · Genuine 4.51 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.344 · CE 6.14 · Genuine 4.51 · WER 0.0% | Target 0.325 · CE 6.26 · Genuine 4.20 · WER 0.0% | Target 0.333 · CE 6.27 · Genuine 4.21 · WER 0.0% | Target 0.324 · CE 6.28 · Genuine 4.49 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen. | Target 0.320 · CE 6.34 · Genuine 4.26 · WER 6.2% | Target 0.318 · CE 6.36 · Genuine 4.26 · WER 0.0% | Target 0.315 · CE 6.33 · Genuine 4.43 · WER 0.0% | Target 0.306 · CE 6.36 · Genuine 4.14 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen. | Target 0.315 · CE 6.39 · Genuine 4.17 · WER 0.0% | Target 0.321 · CE 6.31 · Genuine 4.13 · WER 0.0% | Target 0.328 · CE 6.44 · Genuine 4.55 · WER 0.0% | Target 0.313 · CE 6.48 · Genuine 4.46 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.315 ± 0.016 | 6.152 ± 0.061 | 4.583 ± 0.169 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.312 ± 0.003 | 6.239 ± 0.049 | 4.476 ± 0.154 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.311 ± 0.009 | 6.162 ± 0.039 | 4.558 ± 0.037 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.313 ± 0.007 | 6.168 ± 0.092 | 4.503 ± 0.100 | 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.300 · CE 6.18 · Genuine 4.52 · WER 0.0% | Target 0.309 · CE 6.21 · Genuine 4.39 · WER 0.0% | Target 0.300 · CE 6.16 · Genuine 4.53 · WER 0.0% | Target 0.313 · CE 6.09 · Genuine 4.52 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.303 · CE 6.06 · Genuine 4.51 · WER 0.0% | Target 0.311 · CE 6.19 · Genuine 4.33 · WER 0.0% | Target 0.307 · CE 6.12 · Genuine 4.58 · WER 0.0% | Target 0.303 · CE 6.12 · Genuine 4.40 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.325 · CE 6.17 · Genuine 4.83 · WER 0.0% | Target 0.314 · CE 6.25 · Genuine 4.49 · WER 0.0% | Target 0.320 · CE 6.16 · Genuine 4.52 · WER 0.0% | Target 0.314 · CE 6.16 · Genuine 4.46 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.332 · CE 6.20 · Genuine 4.47 · WER 0.0% | Target 0.315 · CE 6.30 · Genuine 4.68 · WER 0.0% | Target 0.315 · CE 6.21 · Genuine 4.60 · WER 0.0% | Target 0.320 · CE 6.30 · Genuine 4.63 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.317 ± 0.006 | 6.146 ± 0.074 | 4.222 ± 0.057 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.315 ± 0.005 | 6.048 ± 0.098 | 4.197 ± 0.095 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.314 ± 0.004 | 6.108 ± 0.087 | 4.366 ± 0.142 | 8.33% ± 16.67 |
| Rank32 × 1.0 | 4 | 0.315 ± 0.004 | 6.134 ± 0.084 | 4.355 ± 0.160 | 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.311 · CE 6.17 · Genuine 4.18 · WER 0.0% | Target 0.320 · CE 6.17 · Genuine 4.19 · WER 0.0% | Target 0.311 · CE 6.17 · Genuine 4.29 · WER 0.0% | Target 0.313 · CE 6.09 · Genuine 4.21 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.323 · CE 6.04 · Genuine 4.21 · WER 0.0% | Target 0.318 · CE 6.04 · Genuine 4.20 · WER 0.0% | Target 0.318 · CE 6.19 · Genuine 4.23 · WER 0.0% | Target 0.320 · CE 6.25 · Genuine 4.56 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.311 · CE 6.17 · Genuine 4.19 · WER 0.0% | Target 0.315 · CE 5.94 · Genuine 4.08 · WER 0.0% | Target 0.318 · CE 6.01 · Genuine 4.38 · WER 33.3% | Target 0.313 · CE 6.06 · Genuine 4.41 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 Jealousy and Envy. 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 0.322 · CE 6.21 · Genuine 4.31 · WER 0.0% | Target 0.309 · CE 6.05 · Genuine 4.31 · WER 0.0% | Target 0.311 · CE 6.05 · Genuine 4.56 · WER 0.0% | Target 0.313 · CE 6.13 · Genuine 4.25 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.318 ± 0.008 | 5.985 ± 0.057 | 4.418 ± 0.394 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.313 ± 0.004 | 6.245 ± 0.048 | 4.327 ± 0.168 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.312 ± 0.007 | 6.126 ± 0.032 | 4.296 ± 0.348 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.308 ± 0.006 | 6.197 ± 0.054 | 4.352 ± 0.171 | 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta. (Intense Jealousy and Envy. 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo. | Target 0.322 · CE 5.93 · Genuine 4.25 · WER 0.0% | Target 0.309 · CE 6.28 · Genuine 4.34 · WER 0.0% | Target 0.314 · CE 6.12 · Genuine 4.23 · WER 0.0% | Target 0.307 · CE 6.28 · Genuine 4.16 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta. (Intense Jealousy and Envy. 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo. | Target 0.306 · CE 6.06 · Genuine 3.94 · WER 0.0% | Target 0.318 · CE 6.19 · Genuine 4.19 · WER 0.0% | Target 0.304 · CE 6.11 · Genuine 3.86 · WER 0.0% | Target 0.318 · CE 6.17 · Genuine 4.26 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita. (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 0.322 · CE 5.97 · Genuine 4.78 · WER 0.0% | Target 0.311 · CE 6.28 · Genuine 4.56 · WER 0.0% | Target 0.320 · CE 6.10 · Genuine 4.70 · WER 0.0% | Target 0.304 · CE 6.18 · Genuine 4.51 · 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 Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Jealousy and Envy. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita. (Intense Jealousy and Envy. 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 Jealousy and Envy. 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 0.320 · CE 5.98 · Genuine 4.70 · WER 0.0% | Target 0.315 · CE 6.22 · Genuine 4.21 · WER 0.0% | Target 0.309 · CE 6.17 · Genuine 4.39 · WER 0.0% | Target 0.304 · CE 6.16 · Genuine 4.48 · WER 0.0% |