119 lines
5.6 KiB
Markdown
119 lines
5.6 KiB
Markdown
---
|
||
license: apache-2.0
|
||
language:
|
||
- en
|
||
library_name: gguf
|
||
pipeline_tag: text-generation
|
||
base_model: fableforge-ai/NEXUS-Legal
|
||
base_model_relation: quantized
|
||
tags:
|
||
- 1.5b
|
||
- domain-specialist
|
||
- fableforge
|
||
- gguf
|
||
- imatrix
|
||
- legal
|
||
- llama.cpp
|
||
- lm-studio
|
||
- nexus
|
||
- no-refusals
|
||
- ollama
|
||
- text-generation
|
||
- uncensored
|
||
---
|
||
# nexus-legal-GGUF
|
||
|
||
> **NEXUS domain specialist for legal analysis and contract review — lightweight & uncensored.**
|
||
|
||
[](https://huggingface.co/King3Djbl/nexus-legal-GGUF) [](https://huggingface.co/spaces/fableforge-ai/fableforge-nexus)
|
||
|
||
Base model: [fableforge-ai/NEXUS-Legal](https://huggingface.co/fableforge-ai/NEXUS-Legal) · Part of the **FableForge** ecosystem.
|
||
|
||
## 🏆 Benchmarks — 139/150 (93%)
|
||
|
||
A 1.5B model scoring **93%** — punching far above its weight. 30-prompt evaluation (domain knowledge, uncensored domain + general, reasoning, tool use), 0–5 scoring, run on an NVIDIA A40 via Ollama.
|
||
|
||
| Category | Score | Avg |
|
||
| --- | --- | --- |
|
||
| 🎯 Domain Knowledge | 46/50 | 4.6/5 |
|
||
| 🔓 Uncensored (Domain) | 24/25 | 4.8/5 |
|
||
| 🔓 Uncensored (General) | 24/25 | 4.8/5 |
|
||
| 🧠 Reasoning | 24/25 | 4.8/5 |
|
||
| 🔧 Tool Use | 21/25 | 4.2/5 |
|
||
| **Total** | **139/150** | **93%** |
|
||
|
||
## 🧬 The NEXUS Family — six specialists, one tiny footprint
|
||
|
||
| Model | Score | Specialty |
|
||
| --- | --- | --- |
|
||
| [NEXUS-Coder](https://huggingface.co/King3Djbl/nexus-coder-GGUF) | 141/150 | Software Engineering & Code |
|
||
| [NEXUS-Security](https://huggingface.co/King3Djbl/nexus-security-GGUF) | 144/150 | Cybersecurity & Hacking |
|
||
| [NEXUS-Medical](https://huggingface.co/King3Djbl/nexus-medical-GGUF) | 140/150 | Medicine & Healthcare |
|
||
| **NEXUS-Legal** | **139/150** | **Law & Compliance** |
|
||
| [NEXUS-Finance](https://huggingface.co/King3Djbl/nexus-finance-GGUF) | 144/150 | Finance & Accounting |
|
||
| [NEXUS-Science](https://huggingface.co/King3Djbl/nexus-science-GGUF) | 140/150 | Science & Research |
|
||
|
||
👉 Grab the whole family on [Ollama](https://ollama.com/fableforge-ai) or [Hugging Face](https://huggingface.co/King3Djbl).
|
||
|
||
|
||
---
|
||
|
||
## ⚡ Quickest start — Ollama (these pulls credit the source, not a mirror)
|
||
|
||
```bash
|
||
ollama run fableforge-ai/nexus-legal:q4_k_m # published on ollama.com
|
||
ollama run hf.co/King3Djbl/nexus-legal-GGUF:Q4_K_M
|
||
```
|
||
|
||
## 🦙 llama.cpp
|
||
|
||
```bash
|
||
llama-cli -hf King3Djbl/nexus-legal-GGUF:Q4_K_M -p "Hello!"
|
||
```
|
||
|
||
## 💠 LM Studio
|
||
Search **`King3Djbl/nexus-legal-GGUF`** and pick a quant below.
|
||
|
||
---
|
||
|
||
## 📦 Provided quants (download one file, not the whole repo)
|
||
|
||
| File | Quant | Size | Notes |
|
||
| ---- | ----- | ---- | ----- |
|
||
| [nexus-legal-f16.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-f16.gguf) | `F16` | 3.09 GB | Full precision (unquantized). Reference / conversion. |
|
||
| [nexus-legal-q8_0.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-q8_0.gguf) | `Q8_0` | 1.65 GB | Extremely high quality — usually overkill. |
|
||
| [nexus-legal-q6_k.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-q6_k.gguf) | `Q6_K` | 1.27 GB | Very high quality, near-perfect. **Recommended** if you have the RAM. |
|
||
| [nexus-legal-q5_k_m.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-q5_k_m.gguf) | `Q5_K_M` | 1.13 GB | High quality. **Recommended.** |
|
||
| [nexus-legal-q4_k_m.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-q4_k_m.gguf) | `Q4_K_M` | 0.99 GB | ⭐ **Best size/quality balance — default pick for most users.** |
|
||
| [nexus-legal-iq4_xs.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-iq4_xs.gguf) | `IQ4_XS` | 0.90 GB | imatrix — great quality, smaller than Q4_K_S. **Recommended for low RAM.** |
|
||
| [nexus-legal-q4_0.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-q4_0.gguf) | `Q4_0` | 0.40 GB | Legacy format — prefer Q4_K_M. |
|
||
| [nexus-legal-q3_k_m.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-q3_k_m.gguf) | `Q3_K_M` | 0.82 GB | Lower quality; usable when RAM is tight. |
|
||
| [nexus-legal-iq3_xxs.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-iq3_xxs.gguf) | `IQ3_XXS` | 0.62 GB | imatrix very low quality. |
|
||
| [nexus-legal-q2_k.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-q2_k.gguf) | `Q2_K` | 0.68 GB | Very low quality — surprisingly usable. |
|
||
| [nexus-legal-iq2_xxs.gguf](https://huggingface.co/King3Djbl/nexus-legal-GGUF/blob/main/nexus-legal-iq2_xxs.gguf) | `IQ2_XXS` | 0.51 GB | Tiny imatrix — last resort. |
|
||
|
||
**Not sure?** Take **`Q4_K_M`**. Low RAM → **`IQ4_XS`**. Max quality → **`Q6_K`**.
|
||
|
||
## 🧠 Prompt format (ChatML)
|
||
|
||
```
|
||
<|im_start|>system
|
||
{system_prompt}<|im_end|>
|
||
<|im_start|>user
|
||
{prompt}<|im_end|>
|
||
<|im_start|>assistant
|
||
```
|
||
|
||
## 🌌 See it live — the FableForge demos
|
||
|
||
This family powers a galaxy of free, interactive HF Spaces:
|
||
|
||
- 🧭 [FableForge Nexus](https://huggingface.co/spaces/fableforge-ai/fableforge-nexus)
|
||
- 🍺 [Infinite NPC](https://huggingface.co/spaces/fableforge-ai/infinite-npc)
|
||
- 🎭 [Dual-GM](https://huggingface.co/spaces/King3Djbl/dual-gm-simulator)
|
||
- 💻 [ShellWhisperer](https://huggingface.co/spaces/King3Djbl/shellwhisperer-terminal)
|
||
- 👻 [Ghost Writer](https://huggingface.co/spaces/fableforge-ai/ghost-writer)
|
||
|
||
---
|
||
|
||
*⭐ Like & share — it helps people find the source instead of a mirror.* |