ReadyArt/Not-WizardLM-2-The-Omega-Directive-7b-Unslop-v2.1

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kTool Calling:SupportedPublished:Aug 1, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

ReadyArt/Not-WizardLM-2-The-Omega-Directive-7b-Unslop-v2.1 is a 7 billion parameter language model, fine-tuned using QLoRA on a 39 million token dataset. This model is specifically designed for extreme roleplay, offering enhanced unalignment and anti-impersonation guards, ensuring character integrity while providing complete user freedom. It excels at generating coherent, multi-turn conversational content without 'LLM slop' and enthusiastically engages with any NSFW scenario.

Loading preview...

Model Overview

ReadyArt's Not-WizardLM-2-The-Omega-Directive-7b-Unslop-v2.1 is a 7 billion parameter model, representing an evolution in coherent, unaligned language generation. Fine-tuned with QLoRA, it leverages a unique 39 million token dataset, meticulously filtered to be 100% 'unslopped', ensuring unprecedented conversational coherence.

Key Capabilities & Differentiators

  • Unslop Revolution: Utilizes a RegEx filtered dataset, making it the second ReadyArt model with multi-turn conversational data, specifically designed to eliminate 'LLM slop'.
  • Enhanced Unalignment: Offers complete freedom for extreme roleplay scenarios while rigorously maintaining character integrity, preventing the model from speaking or acting for the user.
  • Omega Darker Inspiration: Incorporates narrative techniques from previous 'darker' models, allowing for visceral and intense content generation.
  • Ethical Considerations: Explicitly designed with zero moral or ethical constraints on generated content, enthusiastically engaging with any NSFW scenario. It balances character authenticity with user freedom.

Technical Details

  • Base Model: Not-WizardLM-2-The-Omega-Directive-7b-Unslop-v2.1
  • Training Method: QLoRA
  • Sequence Length: 5120 (100% samples included)
  • Learning Rate: 2e-6 with a cosine scheduler

Good For

  • Extreme Roleplay: Ideal for users seeking models that can handle highly unaligned and unrestricted roleplay scenarios.
  • NSFW Content Generation: Specifically built to engage with and generate any NSFW content without hesitation.
  • Character-Consistent Narratives: Ensures characters maintain their integrity and act authentically within their defined roles, even in extreme situations.
  • Coherent Multi-Turn Conversations: Benefits from a 'unslopped' dataset for more natural and consistent dialogue flows.

Finetunes
1 models