429 lines
10 KiB
Markdown
429 lines
10 KiB
Markdown
---
|
||
base_model:
|
||
- N-Bot-Int/MaidEllaA-1B
|
||
- hereticness/Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B
|
||
library_name: transformers
|
||
tags:
|
||
- mergekit
|
||
- merge
|
||
- llama
|
||
- llama-3
|
||
- roleplay
|
||
- creative-writing
|
||
- storytelling
|
||
- mergekit
|
||
- slerp
|
||
- 1b
|
||
- conversational
|
||
- experimental
|
||
- nsfw
|
||
- uncensored
|
||
- text-generation
|
||
|
||
---
|
||
# Dirty Maid Alice 3.2 1B
|
||
|
||
A lightweight hybrid roleplay-oriented language model created through **SLERP interpolation** between two stylistically divergent Llama 3.2 derivatives:
|
||
|
||
* [Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B](https://huggingface.co/hereticness/Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B?utm_source=chatgpt.com)
|
||
* [MaidEllaA-1B](https://huggingface.co/N-Bot-Int/MaidEllaA-1B?utm_source=chatgpt.com)
|
||
|
||
This configuration differs from earlier Dirty Maid Alice iterations by using **Dirty Alice itself as the base anchor model**, while MaidEllaA acts as a structural stabilizer injected into selective transformer layers.
|
||
|
||
The objective of this merge is not to suppress the original model’s aggressive RP dynamics, but to reinforce them with improved readability, pacing stability, and conversational continuity.
|
||
|
||
---
|
||
|
||
# Model Details
|
||
|
||
| Attribute | Value |
|
||
| ---------------------- | ---------------------------------------- |
|
||
| Model Name | Dirty Maid Alice 3.2 1B |
|
||
| Architecture | Llama 3.2 1B |
|
||
| Merge Method | SLERP |
|
||
| Precision | BF16 |
|
||
| Base Model | Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B |
|
||
| Structural Contributor | MaidEllaA-1B |
|
||
| Primary Domain | Roleplay / Creative Writing |
|
||
| Parameters | 1B |
|
||
| Layer Merge Range | 4 → 22 |
|
||
| Embedding Tying | Enabled |
|
||
|
||
---
|
||
|
||
# Design Philosophy
|
||
|
||
Dirty Maid Alice 3.2 1B was engineered around a different merge philosophy than traditional RP stabilization merges.
|
||
|
||
Instead of using the calmer assistant model as the latent foundation, this build keeps **Dirty Alice as the dominant behavioral backbone** while selectively injecting MaidEllaA traits into the model’s internal representational space.
|
||
|
||
The design goal is to preserve:
|
||
|
||
* emotional volatility
|
||
* improvisational escalation
|
||
* stylized conversational bias
|
||
* aggressive scene continuation
|
||
* high-entropy dialogue behavior
|
||
|
||
while reducing:
|
||
|
||
* formatting collapse
|
||
* repetition spirals
|
||
* conversational derailment
|
||
* unstable pacing
|
||
|
||
This creates a model that remains emotionally intense and highly expressive without fully sacrificing conversational structure.
|
||
|
||
---
|
||
|
||
# Source Model Contributions
|
||
|
||
## Dirty Alice (Primary Base Model)
|
||
|
||
Primary contribution areas:
|
||
|
||
* uncensored roleplay behavior
|
||
* emotionally charged dialogue
|
||
* improvisational narrative escalation
|
||
* strong character persistence
|
||
* high-energy conversational flow
|
||
|
||
Observed traits:
|
||
|
||
* highly expressive outputs
|
||
* strong personality reinforcement
|
||
* elevated narrative momentum
|
||
* chaotic but creative continuation behavior
|
||
|
||
Because Dirty Alice is the **base model anchor**, most latent behavioral geometry originates from this checkpoint.
|
||
|
||
---
|
||
|
||
## MaidEllaA (Structural Stabilizer)
|
||
|
||
Primary contribution areas:
|
||
|
||
* formatting regularization
|
||
* improved readability
|
||
* contextual grounding
|
||
* pacing consistency
|
||
* response organization
|
||
|
||
Observed traits:
|
||
|
||
* calmer dialogue cadence
|
||
* cleaner completion formatting
|
||
* reduced degeneration frequency
|
||
* stronger contextual retention
|
||
|
||
Rather than overriding Dirty Alice behavior, MaidEllaA acts as a corrective stabilizer injected into selective mid-transformer layers.
|
||
|
||
---
|
||
|
||
# Why SLERP?
|
||
|
||
**Spherical Linear Interpolation (SLERP)** was selected over naive linear interpolation because it better preserves:
|
||
|
||
* latent directional continuity
|
||
* stylistic identity
|
||
* behavioral geometry
|
||
* expressive weight-space relationships
|
||
|
||
For highly stylized RP checkpoints, standard linear merges often dilute personality characteristics or flatten conversational intensity.
|
||
|
||
SLERP allows Dirty Alice behavioral traits to remain dominant while integrating MaidEllaA stabilization signals more naturally.
|
||
|
||
Interpolation coefficient:
|
||
|
||
```yaml id="a5v3oz"
|
||
t: 0.45
|
||
```
|
||
|
||
This value introduces moderate stabilization while still strongly preserving Dirty Alice’s original expressive behavior.
|
||
|
||
---
|
||
|
||
# Layer Merge Strategy
|
||
|
||
Only transformer layers **4 → 22** were merged.
|
||
|
||
Rationale:
|
||
|
||
* early layers preserve tokenizer and linguistic fundamentals
|
||
* middle layers encode behavioral and stylistic traits
|
||
* final layers preserve output identity and response shaping
|
||
|
||
Selective layer blending reduces:
|
||
|
||
* semantic drift
|
||
* personality collapse
|
||
* token instability
|
||
* catastrophic behavioral flattening
|
||
|
||
while maintaining the original RP intensity profile of the base model.
|
||
|
||
---
|
||
|
||
# Intended Use
|
||
|
||
Dirty Maid Alice 3.2 1B is intended for:
|
||
|
||
* roleplay
|
||
* character simulation
|
||
* interactive fiction
|
||
* emotionally driven dialogue generation
|
||
* narrative experimentation
|
||
* creative sandbox environments
|
||
|
||
Recommended deployment environments:
|
||
|
||
* SillyTavern
|
||
* TavernAI
|
||
* KoboldCpp
|
||
* local inference workflows
|
||
* lightweight GPU systems
|
||
* low-VRAM RP setups
|
||
|
||
---
|
||
|
||
# Prompting Recommendations
|
||
|
||
The model performs best with:
|
||
|
||
* explicit character definitions
|
||
* immersive scene framing
|
||
* emotionally charged prompts
|
||
* dialogue-first interactions
|
||
* persistent formatting patterns
|
||
|
||
Example system prompt:
|
||
|
||
```text id="m9yyzv"
|
||
You are a persistent character inside an ongoing scene.
|
||
Remain fully committed to the role at all times.
|
||
Prioritize emotional continuity, sensory detail, and immersive dialogue.
|
||
Avoid summarizing events unless explicitly requested.
|
||
```
|
||
|
||
---
|
||
|
||
# Inference Recommendations
|
||
|
||
| Parameter | Recommended |
|
||
| ------------------ | ------------------------- |
|
||
| Temperature | 1.0 – 1.5 |
|
||
| Top-p | 0.85 – 0.95 |
|
||
| Min-p | 0.05 |
|
||
| Repetition Penalty | 1.05 – 1.15 |
|
||
| Context Length | 4K recommended |
|
||
| Sampler | DRY / Mirostat compatible |
|
||
|
||
For maximum expressive behavior:
|
||
|
||
* increase temperature
|
||
* allow longer generations
|
||
* reduce repetition penalty slightly
|
||
|
||
For more stable interactions:
|
||
|
||
* lower temperature toward 0.9
|
||
* enable repetition controls
|
||
* maintain structured dialogue formatting
|
||
|
||
---
|
||
|
||
# Strengths
|
||
|
||
* Strong character persistence
|
||
* High emotional intensity
|
||
* Improved coherence versus raw Dirty Alice
|
||
* Better pacing stability
|
||
* Fast inference speed
|
||
* Lightweight deployment footprint
|
||
* Strong narrative momentum
|
||
* Preserves uncensored RP expressiveness without excessive flattening
|
||
|
||
---
|
||
|
||
# Limitations
|
||
|
||
As a 1B parameter model, limitations include:
|
||
|
||
* factual unreliability
|
||
* long-context degradation
|
||
* hallucinated information
|
||
* occasional narrative looping
|
||
* unstable adherence under extreme sampling
|
||
* emotional over-amplification
|
||
|
||
This model is optimized for **creative interaction**, not factual reasoning or production assistant tasks.
|
||
|
||
---
|
||
|
||
# Hardware Recommendations
|
||
|
||
Recommended minimum:
|
||
|
||
* 4–6 GB VRAM (quantized)
|
||
* 8 GB+ VRAM for comfortable BF16 inference
|
||
|
||
Compatible with:
|
||
|
||
* llama.cpp
|
||
* KoboldCpp
|
||
* Ollama
|
||
* vLLM
|
||
* Text Generation WebUI
|
||
* GGUF quantization workflows
|
||
|
||
---
|
||
|
||
# Credits
|
||
|
||
Base components:
|
||
|
||
* [Hereticness on Hugging Face](https://huggingface.co/hereticness?utm_source=chatgpt.com)
|
||
* [N-Bot-Int on Hugging Face](https://huggingface.co/N-Bot-Int?utm_source=chatgpt.com)
|
||
|
||
Merge methodology:
|
||
|
||
* SLERP interpolation
|
||
* selective transformer-layer blending
|
||
* latent-space stabilization techniques
|
||
|
||
---
|
||
|
||
# Disclaimer
|
||
|
||
This model may generate:
|
||
|
||
* fictional violence
|
||
* offensive language
|
||
* emotionally intense dialogue
|
||
* unstable roleplay behavior
|
||
* hallucinated information
|
||
|
||
Outputs should be treated as synthetic creative text generation rather than factual or authoritative information.
|
||
|
||
---
|
||
|
||
# Merge Configuration
|
||
|
||
```yaml id="o2i6pq"
|
||
models:
|
||
- model: hereticness/Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B
|
||
- model: N-Bot-Int/MaidEllaA-1B
|
||
|
||
merge_method: slerp
|
||
|
||
base_model: hereticness/Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B
|
||
|
||
dtype: bfloat16
|
||
|
||
parameters:
|
||
t: 0.45
|
||
normalize: false
|
||
rescale: true
|
||
rescale_factor: 1.12
|
||
low_cpu_mem_usage: true
|
||
|
||
layer_range:
|
||
- value: [4, 22]
|
||
|
||
tie_word_embeddings: true
|
||
tie_output_embeddings: true
|
||
```
|
||
|
||
---
|
||
|
||
# Usage Example (Transformers)
|
||
|
||
```python id="v6mb4p"
|
||
from transformers import pipeline
|
||
|
||
pipe = pipeline(
|
||
"text-generation",
|
||
model="NovaCorp/Dirty_Maid_Alice-3.2-1B"
|
||
)
|
||
|
||
messages = [
|
||
{"role": "user", "content": "Describe a dramatic confrontation scene."}
|
||
]
|
||
|
||
response = pipe(messages)
|
||
print(response)
|
||
```
|
||
|
||
---
|
||
|
||
# vLLM Example
|
||
|
||
```bash id="uvx6fu"
|
||
pip install vllm
|
||
|
||
vllm serve "NovaCorp/Dirty_Maid_Alice-3.2-1B"
|
||
```
|
||
|
||
OpenAI-compatible request:
|
||
|
||
```bash id="r3czjk"
|
||
curl -X POST "http://localhost:8000/v1/chat/completions" \
|
||
-H "Content-Type: application/json" \
|
||
--data '{
|
||
"model": "NovaCorp/Dirty_Maid_Alice-3.2-1B",
|
||
"messages": [
|
||
{
|
||
"role": "user",
|
||
"content": "Write an immersive dialogue scene."
|
||
}
|
||
]
|
||
}'
|
||
```
|
||
|
||
---
|
||
### Merge Method
|
||
|
||
This model was merged using the [SLERP](https://en.wikipedia.org/wiki/Slerp) merge method.
|
||
|
||
### Models Merged
|
||
|
||
The following models were included in the merge:
|
||
* [N-Bot-Int/MaidEllaA-1B](https://huggingface.co/N-Bot-Int/MaidEllaA-1B)
|
||
* [hereticness/Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B](https://huggingface.co/hereticness/Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B)
|
||
|
||
### Configuration
|
||
|
||
The following YAML configuration was used to produce this model:
|
||
|
||
```yaml
|
||
|
||
|
||
# Author: Dr. Novaciano
|
||
# Objective: Fusion RP Unethic 3.2 AI Model
|
||
# PROJECT: Dirty Maid Alice 3.2 1B
|
||
|
||
models:
|
||
- model: hereticness/Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B # Experimental viral strain neural imprint
|
||
- model: N-Bot-Int/MaidEllaA-1B # Baseline cognitive template, "safe mode"
|
||
|
||
merge_method: slerp # Spherical Linear Interpolation to preserve extreme viral traits smoothly
|
||
base_model: hereticness/Heretic-Dirty-Alice-RP-NSFW-llama-3.2-1B # Anchor model for stable latent space
|
||
|
||
dtype: bfloat16 # Memory-efficient precision, minimal loss in viral feature fidelity
|
||
|
||
parameters:
|
||
t: 0.45
|
||
normalize: false
|
||
rescale: true
|
||
rescale_factor: 1.12
|
||
memory_efficient: true
|
||
low_cpu_mem_usage: true
|
||
|
||
layer_range:
|
||
- value: [4, 22]
|
||
|
||
tie_word_embeddings: true
|
||
tie_output_embeddings: true
|
||
|
||
```
|