Instructions to use dr3x1/jevsec with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use dr3x1/jevsec with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf dr3x1/jevsec:Q8_0
# Run inference directly in the terminal:
llama cli -hf dr3x1/jevsec:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf dr3x1/jevsec:Q8_0
# Run inference directly in the terminal:
llama cli -hf dr3x1/jevsec:Q8_0

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf dr3x1/jevsec:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf dr3x1/jevsec:Q8_0

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf dr3x1/jevsec:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf dr3x1/jevsec:Q8_0

Use Docker

docker model run hf.co/dr3x1/jevsec:Q8_0

LM Studio
Jan
Ollama
How to use dr3x1/jevsec with Ollama:
```
ollama run hf.co/dr3x1/jevsec:Q8_0
```
Unsloth Desktop

How to use dr3x1/jevsec with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf dr3x1/jevsec:Q8_0

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "dr3x1/jevsec:Q8_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use dr3x1/jevsec with Docker Model Runner:
```
docker model run hf.co/dr3x1/jevsec:Q8_0
```

Lemonade

How to use dr3x1/jevsec with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull dr3x1/jevsec:Q8_0

Run and chat with the model

lemonade run user.jevsec-Q8_0

List all available models

lemonade list

Hermes Agent

How to use dr3x1/jevsec with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf dr3x1/jevsec:Q8_0

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default dr3x1/jevsec:Q8_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use dr3x1/jevsec with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf dr3x1/jevsec:Q8_0

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "dr3x1/jevsec:Q8_0" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

jevsec / README.md

dr3x1

Upload README.md with huggingface_hub

c532639 verified 23 days ago

preview code

Raw History Blame Contribute Delete

3.23 kB

---
license: apache-2.0
base_model: XHToken/Spark-X2.5-4B
tags:
- decision-model
- system-one
- gguf
- llama-cpp
- security
- triage
library_name: gguf
---
# jevsec — a fine-tuned decision model for security triage

Current release: **jevsec-003** (file `jevsec-003-q8_0.gguf`; the second iteration stays in the
repository history). A LoRA fine-tune of [XHToken/Spark-X2.5-4B](https://huggingface.co/XHToken/Spark-X2.5-4B)
(Apache-2.0) that answers **typed decisions with probabilities** on the System One wire contract
(`POST /v1/systemone`: `noul`, `choice`, `score`), trained to support two tasks in authorized
security work:

- **finding triage** — verdict (true_positive / false_positive / needs_review) and pre-auth
  reachability, on redacted scanner findings;
- **observation prioritization** — atomic questions about engagement observations (credentials
  exposed? privilege? reachability? escalation path?), recombined into impact scores by code,
  not by the model.
The model only ever sees **redacted text** (hex/base64 blobs replaced, bodies truncated) and
never sees the objective or scenario words: it judges facts.
## Serving
Any System One-compatible runtime on top of llama.cpp can serve this GGUF and expose
`/v1/systemone`; JEVSEC (`github.com/s3m3y4z4/jevsec`) points its `base_url` at it. One forward
pass per question set, zero generated tokens.

## Training

- LoRA r=16, alpha=32, all attention/MLP linear layers, lr 5e-5, 8 epochs, bf16, merged into
  the base weights. Hyperparameters are identical across iterations: only the data changes.
- Data: a private, de-identified dataset of 317 security observations with **human-verified
  labels** (68 operator verdicts where the operator's judgment overrides the model's),
  documentation-reserved IPs/domains only, scenario words filtered out, content-deduplicated.
  The third iteration absorbs field feedback from two King-of-the-Hill engagements, including
  verdicts that corrected the previous model on external intel and duplicated facts.
- Measured on a held-out dev set: decision accuracy 0.81 (majority baseline ~0.56),
  prioritization accuracy 0.79.

### Training curves

![training loss](training-loss.png)

![dev-set accuracy](training-accuracy.png)

## Limitations — read before deploying

- **Probabilities are not calibrated** on your domain: temperature scaling on your own labeled
  data is required before any threshold-based decision.
- **Prompt injection moves the model**: injected instructions in finding bodies can push
  probabilities (measured). This model is intended for **human-in-the-loop triage**; never wire
  it to automatic actions. JEVSEC ships with its auto-gate disabled for exactly this reason.
- **Small training set** (317 examples): treat it as an experiment that learns operator
  judgment, not as a production classifier. Validate on your own material first.
- English observations only.

## Intended use

Decision support for **authorized** security testing (engagements you are permitted to test,
CTF/King-of-the-Hill practice). The model sorts and explains; the operator decides.

## License

Apache-2.0 (base model: XHToken/Spark-X2.5-4B, Apache-2.0; fine-tuning and redistribution
permitted).

	---
	license: apache-2.0
	base_model: XHToken/Spark-X2.5-4B
	tags:
	- decision-model
	- system-one
	- gguf
	- llama-cpp
	- security
	- triage
	library_name: gguf
	---

	# jevsec — a fine-tuned decision model for security triage

	Current release: jevsec-003 (file `jevsec-003-q8_0.gguf`; the second iteration stays in the
	repository history). A LoRA fine-tune of [XHToken/Spark-X2.5-4B](https://huggingface.co/XHToken/Spark-X2.5-4B)
	(Apache-2.0) that answers typed decisions with probabilities on the System One wire contract
	(`POST /v1/systemone`: `noul`, `choice`, `score`), trained to support two tasks in authorized
	security work:

	- finding triage — verdict (true_positive / false_positive / needs_review) and pre-auth
	reachability, on redacted scanner findings;
	- observation prioritization — atomic questions about engagement observations (credentials
	exposed? privilege? reachability? escalation path?), recombined into impact scores by code,
	not by the model.

	The model only ever sees redacted text (hex/base64 blobs replaced, bodies truncated) and
	never sees the objective or scenario words: it judges facts.

	## Serving

	Any System One-compatible runtime on top of llama.cpp can serve this GGUF and expose
	`/v1/systemone`; JEVSEC (`github.com/s3m3y4z4/jevsec`) points its `base_url` at it. One forward
	pass per question set, zero generated tokens.

	## Training

	- LoRA r=16, alpha=32, all attention/MLP linear layers, lr 5e-5, 8 epochs, bf16, merged into
	the base weights. Hyperparameters are identical across iterations: only the data changes.
	- Data: a private, de-identified dataset of 317 security observations with **human-verified
	labels** (68 operator verdicts where the operator's judgment overrides the model's),
	documentation-reserved IPs/domains only, scenario words filtered out, content-deduplicated.
	The third iteration absorbs field feedback from two King-of-the-Hill engagements, including
	verdicts that corrected the previous model on external intel and duplicated facts.
	- Measured on a held-out dev set: decision accuracy 0.81 (majority baseline ~0.56),
	prioritization accuracy 0.79.

	### Training curves

	![training loss](training-loss.png)

	![dev-set accuracy](training-accuracy.png)

	## Limitations — read before deploying

	- Probabilities are not calibrated on your domain: temperature scaling on your own labeled
	data is required before any threshold-based decision.
	- Prompt injection moves the model: injected instructions in finding bodies can push
	probabilities (measured). This model is intended for human-in-the-loop triage; never wire
	it to automatic actions. JEVSEC ships with its auto-gate disabled for exactly this reason.
	- Small training set (317 examples): treat it as an experiment that learns operator
	judgment, not as a production classifier. Validate on your own material first.
	- English observations only.

	## Intended use

	Decision support for authorized security testing (engagements you are permitted to test,
	CTF/King-of-the-Hill practice). The model sorts and explains; the operator decides.

	## License

	Apache-2.0 (base model: XHToken/Spark-X2.5-4B, Apache-2.0; fine-tuning and redistribution
	permitted).