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 5.0.
| Held-out condition | n | Target z↑ | Raw head | CE↑ | Genuine↑ | Voice↑ | WER↓ | Valid | QA |
|---|---|---|---|---|---|---|---|---|---|
| standard | 50 | -0.452 | 0.173 | 6.230 | 4.352 | 0.587 | 0.009 | 1.000 | pass |
| manual | 50 | -0.028 | 0.470 | 5.678 | 4.511 | 0.552 | 0.026 | 0.960 | pass |
| adapter | 50 | -0.140 | 0.392 | 5.760 | 4.520 | 0.542 | 0.020 | 0.980 | pass |
| matched_no_adapter | 50 | -0.088 | 0.428 | 5.787 | 4.565 | 0.568 | 0.016 | 1.000 | pass |
Incremental adapter gain versus same-CFG control: -0.052 z;95% cluster interval [-0.1308398406194371, 0.031035150563198027]. Versus best manual-only: -0.112 z. Manual-only versus standard: 0.424 z.
| Scale | CFG | Target z | CE | Genuine | WER | Provisional QA |
|---|---|---|---|---|---|---|
| 0.0 | 1.0 | -0.449 | 6.178 | 4.540 | 0.000 | True |
| 0.5 | 1.0 | -0.424 | 6.154 | 4.495 | 0.005 | True |
| 0.75 | 1.0 | -0.414 | 6.152 | 4.477 | 0.008 | True |
| 1.0 | 1.0 | -0.441 | 6.142 | 4.463 | 0.003 | True |
| 0.0 | 1.25 | -0.380 | 6.116 | 4.554 | 0.005 | True |
| 0.5 | 1.25 | -0.409 | 6.145 | 4.538 | 0.003 | True |
| 0.75 | 1.25 | -0.324 | 6.115 | 4.529 | 0.003 | True |
| 1.0 | 1.25 | -0.378 | 6.135 | 4.497 | 0.000 | True |
| 0.0 | 1.5 | -0.341 | 6.145 | 4.671 | 0.003 | True |
| 0.5 | 1.5 | -0.342 | 6.150 | 4.590 | 0.008 | True |
| 0.75 | 1.5 | -0.369 | 6.100 | 4.559 | 0.011 | True |
| 1.0 | 1.5 | -0.339 | 6.046 | 4.569 | 0.006 | True |
| 0.0 | 2.0 | -0.232 | 6.051 | 4.622 | 0.003 | True |
| 0.5 | 2.0 | -0.189 | 6.086 | 4.667 | 0.006 | True |
| 0.75 | 2.0 | -0.241 | 6.021 | 4.645 | 0.006 | True |
| 1.0 | 2.0 | -0.348 | 6.028 | 4.587 | 0.008 | True |
| 0.0 | 2.5 | -0.195 | 6.009 | 4.681 | 0.003 | True |
| 0.5 | 2.5 | -0.179 | 5.918 | 4.633 | 0.003 | True |
| 0.75 | 2.5 | -0.113 | 5.895 | 4.623 | 0.006 | True |
| 1.0 | 2.5 | -0.250 | 5.948 | 4.637 | 0.008 | True |
| 0.0 | 3.0 | -0.131 | 5.889 | 4.669 | 0.009 | True |
| 0.5 | 3.0 | -0.213 | 5.921 | 4.666 | 0.003 | True |
| 0.75 | 3.0 | -0.174 | 5.751 | 4.588 | 0.003 | True |
| 1.0 | 3.0 | -0.146 | 5.905 | 4.666 | 0.003 | True |
| 0.0 | 4.0 | -0.086 | 5.800 | 4.632 | 0.013 | True |
| 0.5 | 4.0 | -0.003 | 5.750 | 4.652 | 0.005 | True |
| 0.75 | 4.0 | -0.063 | 5.717 | 4.613 | 0.006 | True |
| 1.0 | 4.0 | -0.144 | 5.705 | 4.614 | 0.000 | True |
| 0.0 | 5.0 | -0.002 | 5.679 | 4.616 | 0.012 | True |
| 0.5 | 5.0 | -0.015 | 5.646 | 4.596 | 0.028 | True |
| 0.75 | 5.0 | -0.132 | 5.557 | 4.538 | 0.003 | True |
| 1.0 | 5.0 | -0.070 | 5.607 | 4.551 | 0.008 | True |
direct_manual_recipe
{
"id": "manual-3a61628184eb",
"general": "Introduce a hesitant, almost fragile texture to the voice. The energy should feel like it's leaking away rather than being actively suppressed.",
"inline": [
"Begin with a slight hesitation, as if the speaker is gathering the courage to utter the words at all.",
"Maintain a breathy quality across the middle phrase, suggesting the effort of maintaining composure is draining.",
"Let the final inflection trail off into an almost inaudible whisper, signifying complete emotional exhaustion."
],
"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 Disappointment. 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: {"de": 47, "nl": 42, "es": 41, "fr": 36, "en": 35}. 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 | 0.313 ± 0.037 | 6.150 ± 0.145 | 4.348 ± 0.159 | 0.28% ± 1.24 |
| Rank32 × 0.5 | 20 | 0.309 ± 0.035 | 6.191 ± 0.137 | 4.313 ± 0.234 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 20 | 0.315 ± 0.030 | 6.137 ± 0.113 | 4.279 ± 0.219 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 20 | 0.319 ± 0.047 | 6.141 ± 0.153 | 4.314 ± 0.218 | 0.31% ± 1.40 |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.293 ± 0.047 | 6.229 ± 0.093 | 4.371 ± 0.129 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.304 ± 0.017 | 6.287 ± 0.071 | 4.220 ± 0.217 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.294 ± 0.031 | 6.174 ± 0.140 | 4.120 ± 0.284 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.295 ± 0.021 | 6.249 ± 0.122 | 4.189 ± 0.312 | 1.56% ± 3.12 |
| 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.284 · CE 6.21 · Genuine 4.34 · WER 0.0% | Target 0.329 · CE 6.32 · Genuine 4.12 · WER 0.0% | Target 0.295 · CE 6.02 · Genuine 3.96 · WER 0.0% | Target 0.320 · CE 6.17 · Genuine 4.13 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.353 · CE 6.11 · Genuine 4.20 · WER 0.0% | Target 0.299 · CE 6.18 · Genuine 3.97 · WER 0.0% | Target 0.332 · CE 6.09 · Genuine 3.80 · WER 0.0% | Target 0.295 · CE 6.12 · Genuine 3.80 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.239 · CE 6.32 · Genuine 4.48 · WER 0.0% | Target 0.295 · CE 6.33 · Genuine 4.46 · WER 0.0% | Target 0.256 · CE 6.28 · Genuine 4.34 · WER 0.0% | Target 0.295 · CE 6.36 · Genuine 4.56 · WER 6.2% |
| 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.296 · CE 6.28 · Genuine 4.46 · WER 0.0% | Target 0.294 · CE 6.31 · Genuine 4.33 · WER 0.0% | Target 0.295 · CE 6.30 · Genuine 4.38 · WER 0.0% | Target 0.269 · CE 6.34 · Genuine 4.26 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.302 ± 0.041 | 6.277 ± 0.176 | 4.391 ± 0.162 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.290 ± 0.048 | 6.332 ± 0.056 | 4.382 ± 0.180 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.312 ± 0.043 | 6.221 ± 0.146 | 4.317 ± 0.075 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.300 ± 0.046 | 6.263 ± 0.160 | 4.339 ± 0.095 | 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.349 · CE 6.17 · Genuine 4.58 · WER 0.0% | Target 0.334 · CE 6.32 · Genuine 4.46 · WER 0.0% | Target 0.357 · CE 6.16 · Genuine 4.35 · WER 0.0% | Target 0.339 · CE 6.24 · Genuine 4.46 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.323 · CE 6.09 · Genuine 4.43 · WER 0.0% | Target 0.328 · CE 6.27 · Genuine 4.60 · WER 0.0% | Target 0.336 · CE 6.05 · Genuine 4.35 · WER 0.0% | Target 0.339 · CE 6.05 · Genuine 4.34 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen. | Target 0.263 · CE 6.39 · Genuine 4.36 · WER 0.0% | Target 0.255 · CE 6.34 · Genuine 4.19 · WER 0.0% | Target 0.290 · CE 6.39 · Genuine 4.36 · WER 0.0% | Target 0.248 · CE 6.38 · Genuine 4.22 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen. | Target 0.272 · CE 6.46 · Genuine 4.19 · WER 0.0% | Target 0.242 · CE 6.40 · Genuine 4.28 · WER 0.0% | Target 0.263 · CE 6.28 · Genuine 4.21 · WER 0.0% | Target 0.274 · CE 6.39 · Genuine 4.33 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.308 ± 0.028 | 6.110 ± 0.096 | 4.376 ± 0.122 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.309 ± 0.039 | 6.191 ± 0.064 | 4.391 ± 0.094 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.311 ± 0.012 | 6.080 ± 0.050 | 4.316 ± 0.167 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.322 ± 0.030 | 6.017 ± 0.075 | 4.418 ± 0.108 | 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.334 · CE 6.21 · Genuine 4.37 · WER 0.0% | Target 0.344 · CE 6.17 · Genuine 4.53 · WER 0.0% | Target 0.320 · CE 6.09 · Genuine 4.29 · WER 0.0% | Target 0.349 · CE 6.01 · Genuine 4.29 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.332 · CE 5.98 · Genuine 4.33 · WER 0.0% | Target 0.340 · CE 6.14 · Genuine 4.34 · WER 0.0% | Target 0.322 · CE 6.13 · Genuine 4.18 · WER 0.0% | Target 0.347 · CE 5.97 · Genuine 4.41 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.284 · CE 6.09 · Genuine 4.55 · WER 0.0% | Target 0.290 · CE 6.28 · Genuine 4.34 · WER 0.0% | Target 0.306 · CE 6.08 · Genuine 4.56 · WER 0.0% | Target 0.295 · CE 5.96 · Genuine 4.56 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.284 · CE 6.16 · Genuine 4.26 · WER 0.0% | Target 0.263 · CE 6.17 · Genuine 4.35 · WER 0.0% | Target 0.298 · CE 6.01 · Genuine 4.24 · WER 0.0% | Target 0.298 · CE 6.13 · Genuine 4.41 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.356 ± 0.020 | 6.102 ± 0.133 | 4.197 ± 0.119 | 1.39% ± 2.78 |
| Rank32 × 0.5 | 4 | 0.343 ± 0.019 | 6.078 ± 0.045 | 4.137 ± 0.187 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.351 ± 0.014 | 6.112 ± 0.072 | 4.274 ± 0.111 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.375 ± 0.064 | 6.064 ± 0.118 | 4.258 ± 0.028 | 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.384 · CE 6.00 · Genuine 4.29 · WER 0.0% | Target 0.330 · CE 6.12 · Genuine 3.90 · WER 0.0% | Target 0.369 · CE 6.12 · Genuine 4.34 · WER 0.0% | Target 0.332 · CE 5.90 · Genuine 4.29 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.353 · CE 5.99 · Genuine 4.25 · WER 0.0% | Target 0.361 · CE 6.01 · Genuine 4.08 · WER 0.0% | Target 0.334 · CE 6.01 · Genuine 4.21 · WER 0.0% | Target 0.470 · CE 6.13 · Genuine 4.26 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.335 · CE 6.15 · Genuine 4.02 · WER 0.0% | Target 0.357 · CE 6.08 · Genuine 4.26 · WER 0.0% | Target 0.351 · CE 6.19 · Genuine 4.16 · WER 0.0% | Target 0.350 · CE 6.16 · Genuine 4.22 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. 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 Disappointment. 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 Disappointment. 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 0.351 · CE 6.27 · Genuine 4.22 · WER 5.6% | Target 0.322 · CE 6.10 · Genuine 4.31 · WER 0.0% | Target 0.349 · CE 6.13 · Genuine 4.39 · WER 0.0% | Target 0.347 · CE 6.07 · Genuine 4.26 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.304 ± 0.016 | 6.035 ± 0.114 | 4.405 ± 0.220 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.301 ± 0.034 | 6.068 ± 0.171 | 4.435 ± 0.368 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.307 ± 0.015 | 6.096 ± 0.117 | 4.368 ± 0.359 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.303 ± 0.020 | 6.112 ± 0.152 | 4.366 ± 0.371 | 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta. (Intense Disappointment. 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo. | Target 0.304 · CE 6.17 · Genuine 4.19 · WER 0.0% | Target 0.349 · CE 6.00 · Genuine 4.48 · WER 0.0% | Target 0.295 · CE 5.95 · Genuine 3.89 · WER 0.0% | Target 0.279 · CE 6.21 · Genuine 4.47 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta. (Intense Disappointment. 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo. | Target 0.282 · CE 6.04 · Genuine 4.24 · WER 0.0% | Target 0.300 · CE 5.86 · Genuine 3.92 · WER 0.0% | Target 0.315 · CE 6.09 · Genuine 4.30 · WER 0.0% | Target 0.318 · CE 5.89 · Genuine 3.82 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita. (Intense Disappointment. 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 Disappointment. 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 0.316 · CE 5.89 · Genuine 4.61 · WER 0.0% | Target 0.267 · CE 6.17 · Genuine 4.56 · WER 0.0% | Target 0.295 · CE 6.24 · Genuine 4.71 · WER 0.0% | Target 0.292 · CE 6.15 · Genuine 4.62 · 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 Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Disappointment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita. (Intense Disappointment. 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 Disappointment. 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 0.315 · CE 6.05 · Genuine 4.58 · WER 0.0% | Target 0.290 · CE 6.25 · Genuine 4.78 · WER 0.0% | Target 0.324 · CE 6.11 · Genuine 4.57 · WER 0.0% | Target 0.322 · CE 6.21 · Genuine 4.56 · WER 0.0% |