Instructions to use SafeAgent/safeagent-7b-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SafeAgent/safeagent-7b-v1 with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="SafeAgent/safeagent-7b-v1",
	filename="safeagent-7b-v1.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use SafeAgent/safeagent-7b-v1 with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf SafeAgent/safeagent-7b-v1
# Run inference directly in the terminal:
llama-cli -hf SafeAgent/safeagent-7b-v1

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf SafeAgent/safeagent-7b-v1
# Run inference directly in the terminal:
llama-cli -hf SafeAgent/safeagent-7b-v1

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 SafeAgent/safeagent-7b-v1
# Run inference directly in the terminal:
./llama-cli -hf SafeAgent/safeagent-7b-v1

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 SafeAgent/safeagent-7b-v1
# Run inference directly in the terminal:
./build/bin/llama-cli -hf SafeAgent/safeagent-7b-v1

Use Docker

docker model run hf.co/SafeAgent/safeagent-7b-v1

LM Studio
Jan
Ollama
How to use SafeAgent/safeagent-7b-v1 with Ollama:
```
ollama run hf.co/SafeAgent/safeagent-7b-v1
```

Unsloth Studio new

How to use SafeAgent/safeagent-7b-v1 with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for SafeAgent/safeagent-7b-v1 to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for SafeAgent/safeagent-7b-v1 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for SafeAgent/safeagent-7b-v1 to start chatting

Docker Model Runner
How to use SafeAgent/safeagent-7b-v1 with Docker Model Runner:
```
docker model run hf.co/SafeAgent/safeagent-7b-v1
```

Lemonade

How to use SafeAgent/safeagent-7b-v1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull SafeAgent/safeagent-7b-v1

Run and chat with the model

lemonade run user.safeagent-7b-v1-{{QUANT_TAG}}

List all available models

lemonade list

safeagent-7b-v1 / README.md

Gomzjaz

Update README.md

8d58d60 verified 26 days ago

preview code

raw

history blame contribute delete

1.77 kB

	---
	language:
	- en
	license: apache-2.0
	tags:
	- safeagent
	- mistral
	- instruction-tuning
	- local-ai
	- self-hosted
	---

	# SafeAgent 7B v1

	The AI model powering [SafeAgent](https://safeagent.dev) ? your personal AI agent that runs entirely on your machine.

	SafeAgent lets you chat with AI, read your emails, search the web, manage GitHub, post to Slack, and automate workflows ? all running locally. Your data never leaves your network. No subscriptions. No cloud dependency.

	## Run It Now

	```bash
	ollama run SafeAgent/safeagent-7b-v1
	```

	## What SafeAgent Can Do With This Model

	- Read and send emails (Gmail, Outlook, Yahoo)
	- Search the web and summarise results
	- Write, fix, and explain code
	- Plan trips, book restaurants, compare products
	- Automate workflows and scheduled tasks
	- Everything stays on your machine -- AES-256 encrypted

	## Install SafeAgent

	```bash
	curl -sL https://www.safeagent.dev/docker-compose.yml -o docker-compose.yml && docker compose up
	```

	Open http://localhost:3000 -- running in 30 seconds.

	## Model Details

	\| Property \| Value \|
	\|----------\|-------\|
	\| Base model \| Mistral 7B v0.1 \|
	\| Fine-tuning method \| QLoRA 4-bit \|
	\| Training data \| 80,000 OpenOrca instruction examples \|
	\| Training time \| 20 hours \|
	\| Final loss \| 0.5564 \|
	\| Format \| GGUF f16 \|

	## Why Local AI?

	\| Cloud AI \| SafeAgent \|
	\|----------\|-----------\|
	\| Your data on their servers \| Everything on your machine \|
	\| $20-100/month \| Free forever \|
	\| Rate limited \| No limits -- your hardware \|
	\| Vendor lock-in \| Open source, always \|
	\| They train on your data \| Your data stays yours \|

	## Links

	- Website: https://safeagent.dev
	- DockerHub: https://hub.docker.com/u/gurmukhs
	- Built by Gurmukh Singh: https://www.linkedin.com/in/gurmukh-singh-38a1a4246/