license: apache-2.0
tags:

• text-generation
• gguf
• nvfp4
• fp4
• qwen3.5
• uncensored
• vision
• multimodal
• genesis
• llama-cpp
  language:
• en
• multilingual
  pipeline_tag: image-text-to-text
  base_model:
• HauhauCS/Qwen3.5-9B-Uncensored-HauhauCS-Aggressive
  base_model_relation: quantized

🌟 Qwen3.5-9B-Uncensored-Genesis-NVFP4-GGUF

This repository contains the NVFP4 (NVIDIA 4-bit Floating Point) quantized GGUF release of Qwen3.5-9B-Uncensored repaired with LuffyTheFox's Genesis calibration algorithm.

• Original Base Model: HauhauCS/Qwen3.5-9B-Uncensored-HauhauCS-Aggressive
• Genesis Tensor Repair: LuffyTheFox
• NVFP4 Quantization: jan1k
• Quantization Format: NVFP4 (NVIDIA FP4) GGUF (optimized for NVIDIA Blackwell, Ada Lovelace, and Ampere GPUs via llama.cpp NVFP4 CUDA kernels)
• Model Architecture: Dense 9B Parameters, hybrid SSM (DeltaNet / Gated Linear Attention) + full attention
• Context Length: 262K native context
• Multimodal (Vision): Supported via mmproj-Qwen3.5-9B-Uncensored-Genesis-BF16.gguf

⚡ What is Genesis?

Genesis is a post-training data regeneration and calibration algorithm for neural networks (LLM) in GGUF format developed by LuffyTheFox over almost half a year of development with AI assistance. It is optimized, architecture-independent, works with any model in GGUF format, and is based on mathematical statistics.

LLM models often suffer from:

• Saturated weights: activations get stuck, gradients vanish, outputs degrade.
• Scale mismatches: one layer's weights are 10× larger than peers for no valid reason.
• Mean drift: weight distributions shifted positive or negative, breaking symmetry assumptions.
• Zero blocks: zero blocks corrupt the signal, turning training into noise amplification.
• Training noise: increases randomness and ruins output quality.

Genesis fixes all of that without retraining — pure numerical surgery on the raw bytes of the file:

1. Stage 1: Scans ssm_conv1d tensors (handling long-context memory) and repairs balance between heads.
2. Stage 2: Scans blocks in chunks (via 3 parameters) and picks the best fit to weight distribution in the tensor, replacing zero blocks without touching learned structure.
3. Stage 3: Scans for noise via custom SVD (excluding token_embd.weight, output.weight, 1D tensors, bias, and norms). Reduces training noise based on the Marchenko–Pastur law while preserving 99% of signal and learned gradient.

⚡ Sampling Recommendations (Genesis SVD Stabilization)

Due to the removal of tensor noise via Marchenko–Pastur SVD noise-gate stabilization, the model is most stable when Top P and Min P are disabled for coding and precise tasks. Only Temperature and Top K should be used for these profiles.

Recommended Parameters:

🚀 Usage Examples (llama-server / llama.cpp)

1. Standard NVFP4 Inference (Thinking ON, 128K Context)

llama-server \
  --host 0.0.0.0 --port 8080 \
  --model Qwen3.5-9B-Uncensored-Genesis-NVFP4.gguf \
  --n-gpu-layers all \
  --flash-attn on \
  --cache-type-k f16 \
  --cache-type-v f16 \
  --ctx-size 131072 \
  --parallel 1 \
  --batch-size 2048 \
  --ubatch-size 512 \
  --temp 0.6 \
  --top-p 1.0 \
  --top-k 20 \
  --min-p 0.0 \
  --seed 42 \
  --presence-penalty 0 \
  --repeat-penalty 1.0 \
  --jinja \
  --chat-template-file chat_template.jinja \
  --reasoning on \
  --reasoning-effort high \
  --reasoning-preserve \
  --reasoning-format deepseek \
  --chat-template-kwargs '{"enable_thinking":true}'

2. Vision (Multimodal) Support

To enable image understanding, pass the multimodal projector:

  --mmproj mmproj-Qwen3.5-9B-Uncensored-Genesis-BF16.gguf

📦 Files in this Repository

🤝 Credits & Attribution

• LuffyTheFox — Genesis post-training signal restoration & calibration algorithm.
• jan1k — NVFP4 quantization.
• HauhauCS — Base uncensored model Qwen3.5-9B-Uncensored-HauhauCS-Aggressive.

⚡ If you like this Genesis release, consider supporting LuffyTheFox via Telegram Tribute or Hipolink.