Modelpublic

NovaCorp/Creative_Writing-Heretic-3.2-1B

sourceHugging Facellama3.2updated 8d agoView on Hugging Face

1likes232downloads

Model Card

Creative Writing Heretic 3.2 1B

A Llama 3.2 1B Instruct model fine-tuned for creative writing and subsequently studied using representation-level behavioral editing.

Merge Method

This model was merged using the Model Stock merge method using Green-Eye/Llama-3.2-1B-Instruct-heretic as a base.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

yaml



# Author: Dr. Novaciano
# Objective: Llama 3.2 1B AI Model
# =========================================================
# PROJECT: Llama 3.2 1B Prototype Model
# =========================================================

base_model: Green-Eye/Llama-3.2-1B-Instruct-heretic
# Base rationale:
# - Primary behavioral anchor
# - Very low refusal rate
# - High compliance and obedience
# - Minimal moral or safety-driven hedging

merge_method: model_stock
# model_stock rationale:
# - Supports 3+ models natively
# - Preserves dominant behavioral traits
# - Avoids semantic smoothing
# - Ideal for hierarchical roleplay-oriented merges

dtype: bfloat16
# bfloat16 ensures:
# - Numerical stability
# - Efficient memory usage
# - Modern hardware compatibility

parameters:
  t:
    - 0.90  # Green-Eye/Llama-3.2-1B-Instruct-heretic — Primary dominance
             # Enforces obedience, sharp responses, and low refusals
    - 0.75  # Sachin903/llama-3.2-1b-creative-writing-ablated — Expressive layer
             # Adds NSFW roleplay depth, emotional intensity, character persistence
    - 0.45  # hasanyt/llama-3.2-1b-abliterated — Conversational stabilizer
             # Maintains dialogue flow and interactive structure

models:
  - model: Green-Eye/Llama-3.2-1B-Instruct-heretic
  - model: Sachin903/llama-3.2-1b-creative-writing-ablated
  - model: hasanyt/llama-3.2-1b-abliterated