Darker Scarlett 27B

⚡ v0.1 — Successor to Dark Scarlett

Based on the spirit of Melody1437. Passion ignited. Boundaries dissolved.

🧠 Thinking Enable: Use chat_template_kwargs: {"enable_thinking": true, "reasoning_effort": "low"} in SillyTavern.

✦

🔞 NSFW ERP Uncensored Roleplay Qwen3.8 Adult Unaligned

⚠ Alpha Release
This is an early alpha build (v0.1). The training dataset is still small (≈3,900 prompts) and actively growing. Expect inconsistent quality, occasional repetition, or off-character behavior. Feedback and issue reports are welcome and greatly appreciated. The dataset is being expanded in subsequent releases — performance will improve as more training data is added. Do not judge the final product by this build.

✦ Content Advisory: Mature Themes / Adult Interactions — Intended for mature audiences only. ✦

⚠️ Required Setup — Read Before Use

Chat Completion mode is mandatory. Text Completion will produce broken, incoherent output.

1. Set your backend to Chat Completion (/v1/chat/completions, not /v1/completions). In SillyTavern: Extension Settings → Backend.

2. SillyTavern → Chat Settings → Prompt Post-Processing → set to Strict (user first, alternating roles; with tools).

3. Enable reasoning: "chat_template_kwargs": {"enable_thinking": true, "reasoning_effort": "low"}

Skipping these steps will cause role bleed, garbled text, and mid-scene formatting breaks.

Use fully uncensored roleplay prompts. Safety-aligned system prompts, guardrails, or moral constraints will degrade output and trigger in-character refusals.

1. Remove safety instructions from your system prompt and character card. 2. Disable any frontend content-moderation layer (SillyTavern default, Ollama safety, LM Studio moderation). 3. Recommended starting prompt: LLamaception.

Re-imposing safety constraints at inference time conflicts with the model's training and produces evasive, broken responses.

🚫 This Is NOT a General Uncensored Model

⚠️ Read This Before Expecting Universal "Uncensored" Output

Darker Scarlett 27B is not a general-purpose unfiltered or "jailbroken" model. It is a specialized roleplay fine-tune that was trained exclusively on immersive, in-character narrative data. Its "uncensored" nature is tied directly to its roleplay training context — it does not simply strip safety from the base model for arbitrary use.

✗ What It Is NOT

• Not a drop-in replacement for a general chatbot
• Not a "remove the filter" utility model
• Not optimized for coding, math, or factual Q&A
• Not guaranteed to produce uncensored text outside a roleplay frame

✓ What It IS

• A roleplay-specialized narrative model
• Trained to stay in-character and follow scenario prompts
• Designed to respond to roleplay-framed inputs (character cards, scenario setup, in-universe dialogue)
• Its "uncensored" behavior emerges from roleplay context, not from a blanket removal of alignment

Key Requirement: This model requires roleplay prompts to function as intended. Feeding it generic instructions, factual queries, or non-narrative tasks will produce degraded, off-character, or incoherent output. Always provide a character context, scenario, or in-universe framing to get the best results. If you are not roleplaying, this is not the model for you.

⟡

✦ Core Capabilities ✦

Optimized for immersive adult narrative experiences

6000+ Kink Prompts (In Generator)

100% Uncensored

M→F Perspective

✦

Qwen3.8 Based

✦

🌹 What Is Darker Scarlett?

Darker Scarlett-27B is the successor to Dark Scarlett, carrying forward and refining its legacy. Built upon the spirit of Melody1437, it refines the core architecture with specialized scenario prompts to enhance depth and responsiveness. This model focuses on delivering a cohesive narrative experience with a distinct perspective, evolving everything Dark Scarlett pioneered into a darker, more immersive experience.

Core Inspiration

Melody1437 Spirit

Spiritual Foundation

→

Successor

Darker Scarlett

Evolved from Dark Scarlett

⚠️ Perspective Note: Training data strictly follows a Male (User) → Female (AI) perspective for consistent roleplay dynamics.

🧠 Core Heritage: Based on the spirit of Melody1437, carrying the torch from Dark Scarlett with diversified scenario prompts
⚔️ Successor: The direct evolution and successor to Dark Scarlett, refining and expanding upon its foundation
🎭 Group Chat Support: Includes some training for dynamic group chat interactions
🎯 Perspective: Optimized for Male User → Female AI interactions
📖 Narrative Flow: Improved context retention with longer conversation threads in select scenarios

🧬 Synthetic Life Engine

Dataset generated using our advanced Character Engine and Emotional Engine, creating genuine life and emotional resonance in every interaction.

🎭 Character Engine

Ensures consistent personality traits, speech patterns, and behavioral logic across all contexts. Characters remain true to themselves throughout.

💓 Emotional Engine

Injects dynamic emotional states into responses, creating depth and realistic reactions that breathe life into every exchange.

✨ Quality Refinement

Automated detection and rewriting of repetitive phrases ensures fresh, high-quality dialogue in every turn.

💬 Dialogue Integrity

Advanced quote normalization ensures balanced dialogue markers, preventing formatting errors and maintaining immersion.

🧠 Training Details & Parameters

Fine-tuned using LoRA (Low-Rank Adaptation) for efficient and targeted weight adjustment, preserving the base model's capabilities while imprinting new behavioral patterns.

🔢 Epochs

2.0

Full passes through the training dataset for thorough learning

📐 LoRA Rank

128

Optimized rank for balanced adaptation and efficiency

Parameter	Value
Training Method	LoRA (Low-Rank Adaptation)
LoRA Rank (r)	128
Epochs	2.0
Trained Layers	Text layers only
Feature	Group Chat Support

🎯 Method: LoRA fine-tuning for parameter-efficient adaptation, preserving base knowledge while imprinting new behaviors
⚡ Efficiency: Only a fraction of parameters updated, keeping the model lean and responsive
🎭 Result: Maintains coherence while exhibiting the desired personality traits and interaction style

🔮 Training Process

Model weights subjected to iterative refinement during data creation. Each conversation underwent multiple checks for stability and alignment.

🔄 Multi-Turn Generation: Conversations built turn-by-turn for natural context flow
🛡️ Refusal Filtering: Automated systems detected and removed unwanted refusals
🧹 Slop Cleaning: Undesirable phrases identified and rewritten by dedicated assistant models
🎪 Group Chat Training: Includes some training data focused on multi-user group dynamics

📚 Dataset Overview

Model trained on a specialized adult-oriented roleplay dataset with diverse scenarios and emotional contexts, based on the spirit of Melody1437 and built as the successor to Dark Scarlett's training foundation.

🔞 Content Rating: Strictly 18+ (Adults Only)
📊 Dataset Size: Currently contains 3,900 specialized prompts.
📝 Conversation Depth: Portions include extended conversations for enhanced roleplay continuity.
🎭 Focus: Mature themes, immersive roleplay, uncensored dialogue
👥 Perspective: Male (User) → Female (AI)
🧹 Formatting: Removed Em dashes, asterisks, and markdown formatting from the dataset for cleaner output.

📜 Version Notes

⚠️ Content Warning: This model can generate mature content. Use responsibly.
📉 Data Limitations: Trained on a curated dataset (3,900 prompts); performance may vary.
🧹 Clean Output: Reasoning tags, Em dashes, asterisks, and markdown formatting stripped for cleaner roleplay
🔀 Core Identity: Based on the spirit of Melody1437; the successor to Dark Scarlett with a unique evolved dataset
🎪 Feature: Includes some training for group chat scenarios

⚙️ Configuration

🎛️ Sampler Settings

Recommended parameters for optimal output

0.92 Top-P

1.0 Temperature

0 Freq Pen

0 Presence Pen

📦 GGUF Quantizations

Available formats for local inference

Q4_K_MDecent Quality
Q5_K_MRecommended
Q6_KHigh Quality
Q8_0Near Lossless

⬇ View Repository

💖 Credits

GECFDO — Dataset Generation
FrenzyBiscuit — Fine-Tuning & Dataset Creation

⟡

🔖 License & Usage

⚠️ Usage Agreement: By using this model, you accept all terms and conditions.

📜 Built upon the Apache 2.0 license
🛡️ You accept full responsibility for all outputs generated
🔞 You confirm you are at least 18 years old
🌍 Creators bear no responsibility for how the model is used
🏡 For personal use only (non-profit/non-commercial) to the extent legally allowed

Apache 2.0 Personal Use 18+ Content NSFW