---
license: apache-2.0
language:
  - en
base_model: Qwen/Qwen2.5-0.5B
tags:
  - shell
  - commands
  - terminal
  - devops
  - system-admin
  - uncensored
  - fable5
  - ollama
  - lightweight
pipeline_tag: text-generation
---

# ShellWhisperer-1.5B — Ultra-Fast Shell Command Assistant

<div align="center">

**734+ downloads · 986MB · 20+ tok/s · Runs on anything**

[![Ollama](https://img.shields.io/badge/🦙-Ollama-black)](https://ollama.com/FableForge-AI/shellwhisperer)
[![Size](https://img.shields.io/badge/size-986MB-success)](#quantizations)
[![License](https://img.shields.io/badge/license-Apache--2.0-blue)](#license)

</div>

---

## What Is This?

ShellWhisperer is a tiny but powerful 1.5B parameter model fine-tuned specifically for shell command prediction and system administration. Based on Qwen2.5-0.5B, it's ultra-lightweight and runs on virtually any device — phones, Raspberry Pi, old laptops.

## Quick Start

### Ollama
```bash
ollama run FableForge-AI/shellwhisperer
```

### llama.cpp
```bash
./llama-cli --model shellwhisperer-1.5b-Q4_K_M.gguf --prompt "find all files larger than 100MB"
```

---

## Quantizations

| File | Size | Best For |
|------|------|----------|
| `shellwhisperer-1.5b-Q4_K_M.gguf` | 940MB | **Recommended** |
| `shellwhisperer-1.5b-Q5_K_M.gguf` | 1.0GB | High quality |
| `shellwhisperer-1.5b-Q8_0.gguf` | 1.5GB | Max quality |
| `shellwhisperer-1.5b-f16.gguf` | 2.9GB | Full precision |

### Hardware Requirements

| Hardware | Can Run? | Speed |
|----------|----------|-------|
| Any phone (2GB+ RAM) | ✅ | ~20 tok/s |
| Raspberry Pi Zero | ✅ | ~15 tok/s |
| Old laptop | ✅ | ~25 tok/s |
| RTX 3060+ | ✅ Full GPU | ~40 tok/s |
| M1 Mac | ✅ Full GPU | ~35 tok/s |

---

## Capabilities

### Shell Command Prediction
```
User: Find all files modified in last 7 days
ShellWhisperer: find / -type f -mtime -7 2>/dev/null | head -20
```

### System Administration
```
User: Check disk usage of all mounted drives
ShellWhisperer: df -h | grep -v tmpfs | sort -k5 -hr
```

### DevOps
```
User: List all running Docker containers with ports
ShellWhisperer: docker ps --format "table {{.Names}}\t{{.Ports}}\t{{.Status}}"
```

### Git Operations
```
User: Undo last commit but keep changes
ShellWhisperer: git reset --soft HEAD~1
```

---

## Training Details

| Parameter | Value |
|-----------|-------|
| Base Model | Qwen2.5-0.5B |
| Training | Shell command distillation from Fable5 traces |
| Context Window | 32K tokens |
| License | Apache 2.0 |

---

## FableForge Ecosystem

| Model | Size | Best For |
|-------|------|----------|
| **ShellWhisperer** | **986MB** | **Shell commands, ultra-fast** |
| Mythos 9B | 5.0GB | All-rounder uncensored |
| Enhanced | 5.0GB | Perfect bypass + tools |
| Unhinged | 5.0GB | Max speed, zero filter |
| ReasonCritic-7B | 3.1-16GB | Reasoning + phone |

---

## License

Apache 2.0 — commercial use allowed, no restrictions.

---

<div align="center">

⭐ [GitHub](https://github.com/FableForge-AI) · 📦 [Ollama](https://ollama.com/FableForge-AI) · 🤗 [HuggingFace](https://huggingface.co/fableforge-ai)

</div>