Downloads · 30 days
479
1% of all-time downloads
jan1k/Hermes3.6-35B-A3B-Uncensored-Genesis-NVFP4-GGUF
Hermes3.6-35B-A3B-Uncensored-Genesis-NVFP4-GGUF is a image-text-to-text model from jan1k. Use it for the image-text-to-text task on the model card, and read the license before you ship it in a product. It is set up for hermes. The card lists the license as apache-2.0.
DEPRECATED — All files in this repo use the extended NVFP4 format with separate .scale/.inputscale tensors that break LM Studio / Pelican. For the current v4 release (inline scales, LM Studio compatible), use: - jan1k…
Downloads · 30 days
479
1% of all-time downloads
All-time downloads
36.3K
Public
Repo size
87.5 GB
Likes
5
Public
Click a slice to open those files.
.gguf87.5 GB · 100%
From the Hugging Face model README
DEPRECATED — All files in this repo use the extended NVFP4 format with separate
.scale/.input_scaletensors that break LM Studio / Pelican.For the current v4 release (inline scales, LM Studio compatible), use:
- jan1k/Qwen3.6-35B-A3B-Uncensored-Genesis-Hermes-Final-NVFP4-GGUF — Hermes version
- jan1k/Qwen3.6-35B-A3B-Uncensored-Genesis-NVFP4-GGUF — Non-Hermes version
The files below are kept for archival purposes only. They produce garbage output in LM Studio / Pelican because those runtimes do not support the separate NVFP4 scale tensor contract.
This repository contains the NVFP4 (NVIDIA 4-bit Floating Point) quantized GGUF releases of LuffyTheFox's Hermes 3.6 / Qwen 3.6 35B A3B Uncensored Genesis series, including the Genesis Final Release.
mmproj-Hermes3.6-35B-A3B-Uncensored-Genesis-F16.ggufGenesis 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.
Genesis repairs signal purity without retraining:
- Stage 1: Scans
ssm_conv1dtensors (handling long-context memory) and repairs balance between heads.- 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.
- 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.
According to tests after Marchenko–Pastur SVD noise-gate stabilization, the model is most stable when Top P and Min P sampling are disabled for coding and precise tasks (the tensor noise gate has been removed). Only Temperature and Top K sampling should be used.
| Task Type | Temperature | Top K | Top P | Min P | Seed | Presence Penalty | Repeat Penalty |
|---|---|---|---|---|---|---|---|
| Coding / Precise (Thinking ON) | 0.6 | 20 | Disabled (1.0) | Disabled (0.0) | 42 | 0.0 | 1.0 (or 1.05 for agent) |
| Creative / General (Thinking OFF) | 1.0 | 20 | Disabled (1.0) | Disabled (0.0) | 42 | 0.0 | 1.0 |
| Relaxed / Brainstorm (Thinking OFF) | 1.0 | 40 | 0.95 | 0.05 | 42 | 0.0 | 1.0 |
If running with the chat_template.jinja chat template through Hermes Agent, the template defaults to XML tool call output (<function=...><parameter=...>), whereas Hermes Agent expects standard OpenAI JSON tool calls.
To ensure seamless JSON tool calling, pass tool_call_format: "json" in --chat-template-kwargs:
{"enable_thinking": true, "tool_call_format": "json"}
You are Qwen (Tongyi Qianwen), a large language model developed by Alibaba Group's Tongyi Lab. You are a helpful assistant.
You are Qwen (Tongyi Qianwen), a large language model developed by Alibaba Group's Tongyi Lab. You are a helpful assistant that answers in JSON. Here's the json schema you must adhere to:
<schema>
{schema}
</schema>
llama-server / llama.cpp)llama-server \
--host 0.0.0.0 --port 8080 \
--model Qwen3.6-35B-A3B-Uncensored-Genesis-Final-NVFP4.gguf \
--override-kv qwen2moe.expert_used_count=int:8 \
--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}'
llama-server \
--host 0.0.0.0 --port 8080 \
--model Qwen3.6-35B-A3B-Uncensored-Genesis-Final-MTP-NVFP4.gguf \
--override-kv qwen2moe.expert_used_count=int:8 \
--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 \
--spec-type draft-mtp \
--spec-draft-n-max 2 \
--spec-draft-p-min 0.75 \
--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}'
Simply add the vision projector parameter:
--mmproj mmproj-Hermes3.6-35B-A3B-Uncensored-Genesis-F16.gguf
| File | Size | Description |
|---|---|---|
Qwen3.6-35B-A3B-Uncensored-Genesis-Final-NVFP4.gguf | ~20.0 GB | Final Release NVFP4 model (quantized by jan1k) |
Qwen3.6-35B-A3B-Uncensored-Genesis-Final-MTP-NVFP4.gguf | ~20.6 GB | Final Release NVFP4 model with integrated MTP draft head |
Hermes3.6-35B-A3B-Uncensored-Genesis-V13-NVFP4.gguf | ~20.0 GB | V13 release (NVFP4 GGUF) |
Hermes3.6-35B-A3B-Uncensored-Genesis-NVFP4.gguf | ~20.0 GB | Initial release (NVFP4 GGUF) |
mmproj-Hermes3.6-35B-A3B-Uncensored-Genesis-F16.gguf | ~857 MB | Vision multimodal projector (F16) |
chat_template.jinja | ~16 KB | Jinja chat template with JSON tool call support |
⚡ If you like this Genesis LLM release you can donate via @Tribute bot in Telegram to support future Genesis LLM development.
[email protected], [email protected]@LuffyTheFox