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 | 3.113 | 1.923 | 6.257 | 4.338 | 0.560 | 0.008 | 1.000 | pass |
| manual | 50 | 3.859 | 2.299 | 6.276 | 4.290 | 0.557 | 0.032 | 0.960 | pass |
| adapter | 50 | 3.879 | 2.309 | 6.253 | 4.206 | 0.505 | 0.018 | 1.000 | fail |
| matched_no_adapter | 50 | 3.868 | 2.303 | 6.281 | 4.221 | 0.546 | 0.008 | 1.000 | pass |
Incremental adapter gain versus same-CFG control: 0.011 z;95% cluster interval [-0.05968745626300729, 0.09624999334517437]. Versus best manual-only: 0.020 z. Manual-only versus standard: 0.746 z.
| Scale | CFG | Target z | CE | Genuine | WER | Provisional QA |
|---|---|---|---|---|---|---|
| 0.0 | 1.0 | 3.427 | 6.226 | 4.332 | 0.006 | True |
| 0.5 | 1.0 | 3.624 | 6.205 | 4.308 | 0.005 | True |
| 0.75 | 1.0 | 3.460 | 6.202 | 4.270 | 0.003 | True |
| 1.0 | 1.0 | 3.643 | 6.220 | 4.343 | 0.000 | True |
| 0.0 | 1.25 | 3.569 | 6.189 | 4.257 | 0.000 | True |
| 0.5 | 1.25 | 3.542 | 6.246 | 4.311 | 0.003 | True |
| 0.75 | 1.25 | 3.518 | 6.199 | 4.294 | 0.000 | True |
| 1.0 | 1.25 | 3.618 | 6.212 | 4.304 | 0.000 | True |
| 0.0 | 1.5 | 3.537 | 6.219 | 4.319 | 0.003 | True |
| 0.5 | 1.5 | 3.608 | 6.247 | 4.327 | 0.000 | True |
| 0.75 | 1.5 | 3.608 | 6.221 | 4.332 | 0.000 | True |
| 1.0 | 1.5 | 3.691 | 6.231 | 4.270 | 0.000 | True |
| 0.0 | 2.0 | 3.710 | 6.222 | 4.258 | 0.000 | True |
| 0.5 | 2.0 | 3.697 | 6.241 | 4.253 | 0.003 | True |
| 0.75 | 2.0 | 3.778 | 6.203 | 4.280 | 0.003 | True |
| 1.0 | 2.0 | 3.686 | 6.200 | 4.238 | 0.003 | True |
| 0.0 | 2.5 | 3.826 | 6.230 | 4.307 | 0.006 | True |
| 0.5 | 2.5 | 3.740 | 6.234 | 4.245 | 0.003 | True |
| 0.75 | 2.5 | 3.836 | 6.228 | 4.217 | 0.000 | True |
| 1.0 | 2.5 | 3.612 | 6.238 | 4.302 | 0.000 | True |
| 0.0 | 3.0 | 3.906 | 6.235 | 4.225 | 0.003 | True |
| 0.5 | 3.0 | 3.655 | 6.230 | 4.231 | 0.003 | True |
| 0.75 | 3.0 | 3.959 | 6.203 | 4.236 | 0.009 | True |
| 1.0 | 3.0 | 3.905 | 6.219 | 4.276 | 0.000 | True |
| 0.0 | 4.0 | 4.041 | 6.199 | 4.255 | 0.000 | True |
| 0.5 | 4.0 | 3.966 | 6.209 | 4.213 | 0.000 | True |
| 0.75 | 4.0 | 3.854 | 6.234 | 4.239 | 0.003 | True |
| 1.0 | 4.0 | 4.012 | 6.157 | 4.227 | 0.006 | True |
| 0.0 | 5.0 | 4.019 | 6.163 | 4.222 | 0.006 | True |
| 0.5 | 5.0 | 4.005 | 6.185 | 4.175 | 0.006 | True |
| 0.75 | 5.0 | 4.120 | 6.192 | 4.235 | 0.053 | True |
| 1.0 | 5.0 | 4.017 | 6.200 | 4.144 | 0.050 | False |
direct_manual_recipe
{
"id": "manual-e34e2073bf7b",
"general": "Treat the delivery as a shared, comfortable secret; intimate, low-key, and deeply resonant.",
"inline": [
"Start with a near-whisper quality that immediately builds into a rich, comforting mid-range tone, implying shared tranquility.",
"Employ a very slight, almost imperceptible vocal fry on key consonants to enhance the sense of settled, deep comfort.",
"Conclude the speech with a sustained, resonant tone that lingers slightly beyond the last word, like a warm echo."
],
"cfg": 3,
"recipe": "P09",
"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 Contentment. 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": 16, "de": 23, "en": 31, "fr": 13, "nl": 2}. 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 | 0.982 ± 0.055 | 6.208 ± 0.130 | 4.276 ± 0.185 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 20 | 0.970 ± 0.086 | 6.156 ± 0.140 | 4.360 ± 0.230 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 20 | 0.981 ± 0.049 | 6.173 ± 0.138 | 4.387 ± 0.250 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 20 | 0.973 ± 0.069 | 6.131 ± 0.173 | 4.301 ± 0.229 | 0.00% ± 0.00 |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 1.000 ± 0.000 | 6.277 ± 0.157 | 4.139 ± 0.320 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 1.000 ± 0.000 | 6.330 ± 0.048 | 4.316 ± 0.250 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 1.000 ± 0.000 | 6.267 ± 0.051 | 4.115 ± 0.149 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 1.000 ± 0.000 | 6.247 ± 0.133 | 4.245 ± 0.324 | 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 0.999 · CE 6.06 · Genuine 3.74 · WER 0.0% | Target 1.000 · CE 6.28 · Genuine 4.11 · WER 0.0% | Target 1.000 · CE 6.25 · Genuine 3.98 · WER 0.0% | Target 1.000 · CE 6.09 · Genuine 3.79 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 1.000 · CE 6.27 · Genuine 4.02 · WER 0.0% | Target 1.000 · CE 6.31 · Genuine 4.09 · WER 0.0% | Target 1.000 · CE 6.22 · Genuine 3.99 · WER 0.0% | Target 1.000 · CE 6.24 · Genuine 4.35 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 1.000 · CE 6.43 · Genuine 4.43 · WER 0.0% | Target 1.000 · CE 6.35 · Genuine 4.55 · WER 0.0% | Target 1.000 · CE 6.34 · Genuine 4.26 · WER 0.0% | Target 1.000 · CE 6.24 · Genuine 4.28 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 1.000 · CE 6.35 · Genuine 4.37 · WER 0.0% | Target 1.000 · CE 6.39 · Genuine 4.52 · WER 0.0% | Target 1.000 · CE 6.25 · Genuine 4.23 · WER 0.0% | Target 1.000 · CE 6.42 · Genuine 4.56 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.995 ± 0.008 | 6.356 ± 0.096 | 4.203 ± 0.076 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.996 ± 0.004 | 6.251 ± 0.136 | 4.343 ± 0.147 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.999 ± 0.001 | 6.303 ± 0.098 | 4.341 ± 0.264 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.992 ± 0.010 | 6.277 ± 0.104 | 4.327 ± 0.096 | 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 0.997 · CE 6.39 · Genuine 4.17 · WER 0.0% | Target 0.990 · CE 6.14 · Genuine 4.41 · WER 0.0% | Target 0.997 · CE 6.30 · Genuine 4.60 · WER 0.0% | Target 0.990 · CE 6.14 · Genuine 4.39 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 0.983 · CE 6.21 · Genuine 4.23 · WER 0.0% | Target 0.996 · CE 6.13 · Genuine 4.52 · WER 0.0% | Target 0.999 · CE 6.17 · Genuine 4.46 · WER 0.0% | Target 0.979 · CE 6.25 · Genuine 4.40 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen. | Target 1.000 · CE 6.41 · Genuine 4.29 · WER 0.0% | Target 0.999 · CE 6.32 · Genuine 4.24 · WER 0.0% | Target 1.000 · CE 6.38 · Genuine 4.33 · WER 0.0% | Target 1.000 · CE 6.38 · Genuine 4.19 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen. | Target 0.999 · CE 6.41 · Genuine 4.12 · WER 0.0% | Target 1.000 · CE 6.41 · Genuine 4.21 · WER 0.0% | Target 1.000 · CE 6.36 · Genuine 3.98 · WER 0.0% | Target 0.999 · CE 6.33 · Genuine 4.32 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.997 ± 0.006 | 6.172 ± 0.082 | 4.427 ± 0.091 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.999 ± 0.001 | 6.086 ± 0.028 | 4.470 ± 0.350 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.999 ± 0.002 | 6.156 ± 0.057 | 4.652 ± 0.211 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.993 ± 0.013 | 6.080 ± 0.029 | 4.515 ± 0.194 | 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 1.000 · CE 6.18 · Genuine 4.41 · WER 0.0% | Target 0.998 · CE 6.09 · Genuine 4.39 · WER 0.0% | Target 0.997 · CE 6.19 · Genuine 4.56 · WER 0.0% | Target 0.974 · CE 6.05 · Genuine 4.61 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 0.989 · CE 6.06 · Genuine 4.43 · WER 0.0% | Target 1.000 · CE 6.10 · Genuine 4.04 · WER 0.0% | Target 1.000 · CE 6.17 · Genuine 4.41 · WER 0.0% | Target 1.000 · CE 6.09 · Genuine 4.22 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 1.000 · CE 6.18 · Genuine 4.33 · WER 0.0% | Target 1.000 · CE 6.11 · Genuine 4.56 · WER 0.0% | Target 1.000 · CE 6.19 · Genuine 4.76 · WER 0.0% | Target 1.000 · CE 6.12 · Genuine 4.60 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 1.000 · CE 6.26 · Genuine 4.55 · WER 0.0% | Target 1.000 · CE 6.05 · Genuine 4.88 · WER 0.0% | Target 1.000 · CE 6.07 · Genuine 4.88 · WER 0.0% | Target 1.000 · CE 6.06 · Genuine 4.63 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 0.920 ± 0.112 | 6.106 ± 0.064 | 4.289 ± 0.065 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.858 ± 0.161 | 6.050 ± 0.096 | 4.277 ± 0.166 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.913 ± 0.086 | 6.076 ± 0.188 | 4.343 ± 0.053 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 0.881 ± 0.126 | 6.132 ± 0.073 | 4.245 ± 0.084 | 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 0.761 · CE 6.09 · Genuine 4.34 · WER 0.0% | Target 0.693 · CE 6.04 · Genuine 4.14 · WER 0.0% | Target 0.901 · CE 6.24 · Genuine 4.36 · WER 0.0% | Target 0.737 · CE 6.11 · Genuine 4.22 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 0.924 · CE 6.05 · Genuine 4.34 · WER 0.0% | Target 0.747 · CE 5.92 · Genuine 4.12 · WER 0.0% | Target 0.796 · CE 5.82 · Genuine 4.41 · WER 0.0% | Target 0.813 · CE 6.18 · Genuine 4.19 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 0.998 · CE 6.09 · Genuine 4.22 · WER 0.0% | Target 0.999 · CE 6.11 · Genuine 4.44 · WER 0.0% | Target 0.978 · CE 6.06 · Genuine 4.33 · WER 0.0% | Target 0.979 · CE 6.20 · Genuine 4.21 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. 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 Contentment. 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 Contentment. 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 0.998 · CE 6.20 · Genuine 4.25 · WER 0.0% | Target 0.993 · CE 6.13 · Genuine 4.39 · WER 0.0% | Target 0.977 · CE 6.19 · Genuine 4.28 · WER 0.0% | Target 0.994 · CE 6.04 · Genuine 4.37 · WER 0.0% |
Fixed speaker reference:
| Condition | n | Target score ↑ | Content Enjoyment ↑ | Genuineness ↑ | WER ↓ |
|---|---|---|---|---|---|
| Merged base, no bucket adapter | 4 | 1.000 ± 0.000 | 6.128 ± 0.046 | 4.324 ± 0.170 | 0.00% ± 0.00 |
| Rank32 × 0.5 | 4 | 0.999 ± 0.003 | 6.060 ± 0.084 | 4.394 ± 0.267 | 0.00% ± 0.00 |
| Rank32 × 0.75 | 4 | 0.996 ± 0.006 | 6.062 ± 0.083 | 4.484 ± 0.215 | 0.00% ± 0.00 |
| Rank32 × 1.0 | 4 | 1.000 ± 0.000 | 5.921 ± 0.217 | 4.173 ± 0.283 | 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta. (Intense Contentment. 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo. | Target 1.000 · CE 6.13 · Genuine 4.26 · WER 0.0% | Target 1.000 · CE 6.09 · Genuine 4.27 · WER 0.0% | Target 0.996 · CE 6.10 · Genuine 4.23 · WER 0.0% | Target 1.000 · CE 5.63 · Genuine 3.96 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta. (Intense Contentment. 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo. | Target 1.000 · CE 6.07 · Genuine 4.28 · WER 0.0% | Target 0.995 · CE 5.94 · Genuine 4.08 · WER 0.0% | Target 0.987 · CE 6.16 · Genuine 4.53 · WER 0.0% | Target 1.000 · CE 5.90 · Genuine 3.92 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita. (Intense Contentment. 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 Contentment. 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 1.000 · CE 6.13 · Genuine 4.18 · WER 0.0% | Target 1.000 · CE 6.09 · Genuine 4.66 · WER 0.0% | Target 1.000 · CE 5.97 · Genuine 4.43 · WER 0.0% | Target 1.000 · CE 6.14 · 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 Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word. SCRIPT: (Intense Contentment. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita. (Intense Contentment. 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 Contentment. 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 1.000 · CE 6.18 · Genuine 4.57 · WER 0.0% | Target 1.000 · CE 6.12 · Genuine 4.57 · WER 0.0% | Target 1.000 · CE 6.01 · Genuine 4.75 · WER 0.0% | Target 1.000 · CE 6.01 · Genuine 4.31 · WER 0.0% |