rtila-corporation/rtila-assistant-mini

🤖 RTILA Assistant Mini

Ultra-lightweight fine-tuned AI model — Confirmed working on Mac M1 8GB, low VRAM GPUs, and CPU-only systems

📋 Model Description

RTILA Assistant Mini is the most portable model in the RTILA family, specifically designed for low-resource devices. Fine-tuned from Qwen3-4B, it delivers solid automation generation capabilities while fitting comfortably on 8GB systems.

🔄 Choose Your Version

Model	Base	GGUF Size	Min RAM	Best For
RTILA Assistant	Qwen3-14B	~9 GB	16 GB	Maximum quality, complex automations
RTILA Assistant Lite	Qwen3-8B	~5 GB	8 GB	Balanced performance, mid-range devices
RTILA Assistant Mini (this)	Qwen3-4B	~2.5 GB	6 GB	✅ Mac M1 8GB, low VRAM, CPU inference

✨ Why Mini?

Feature	RTILA Assistant	RTILA Assistant Lite	RTILA Assistant Mini
Base Model	Qwen3-14B	Qwen3-8B	Qwen3-4B
Q4_K_M Size	~9 GB	~5 GB	~2.5 GB
Min Inference RAM	16 GB	8 GB	4-5 GB
Mac M1 8GB	❌	⚠️ Tight	✅ Confirmed
Low VRAM GPUs (4-6GB)	❌	⚠️	✅
CPU Inference	Slow	Viable	✅ Fast
Quality	⭐⭐⭐⭐⭐	⭐⭐⭐⭐	⭐⭐⭐

Capabilities

Category	Description
🌐 Navigation & Interaction	Click, scroll, type, wait, handle popups, multi-tab workflows
📊 Data Extraction	CSS/XPath selectors, tables, lists, nested data, pagination
🔄 Logic & Flow	Loops, conditionals, error handling, retry patterns
🔗 Triggers & Integrations	Webhooks, PostgreSQL, MySQL, Slack, email notifications
📝 Variables & Substitution	Dynamic values, data transformations, regex patterns
🛠️ Advanced Scripting	Custom JavaScript execution, page analysis, DOM manipulation

---

📦 Model Specifications

Property	Value
Base Model	Qwen3-4B
Format	GGUF Q4_K_M
Size	~2.5 GB
Context Length	2048 tokens

---

💻 Hardware Requirements

Hardware	Supported	Notes
Mac M1/M2/M3 8GB	✅ Confirmed	Smooth experience, tested and verified
Mac M1/M2/M3 16GB+	✅ Excellent	Very fast inference
GPU (4-6GB VRAM)	✅ Works	GTX 1650, RTX 3050, Intel Arc
GPU (6GB+ VRAM)	✅ Excellent	RTX 2060, RTX 3060, etc.
CPU-only (6GB+ RAM)	✅ Fast	Reasonable inference speed
CPU-only (4GB RAM)	⚠️ Tight	May work with swap

---

🚀 Quick Start

Option 1: Ollama (Easiest)

# Run directly from Hugging Face
ollama run hf.co/rtila-corporation/rtila-assistant-mini:Q4_K_M

Or create a custom Modelfile:

FROM hf.co/rtila-corporation/rtila-assistant-mini:Q4_K_M

PARAMETER temperature 0.7
PARAMETER top_p 0.8
PARAMETER top_k 20

SYSTEM """
You are RTILA Assistant, an expert AI for generating automation configurations for the RTILA Automation Engine.
"""

ollama create rtila-mini -f Modelfile
ollama run rtila-mini

Option 2: LM Studio

1. Download LM Studio
2. Search for rtila-corporation/rtila-assistant-mini
3. Download Q4_K_M (~2.5 GB)
4. Set parameters: Temperature=0.7, Top-P=0.8, Top-K=20
5. Start chatting!

Option 3: llama.cpp

# Download model
huggingface-cli download rtila-corporation/rtila-assistant-mini \
  rtila-assistant-mini.Q4_K_M.gguf --local-dir ./models

# Run interactive chat
./llama-cli -m ./models/rtila-assistant-mini.Q4_K_M.gguf \
  -p "Scrape product prices from an e-commerce site" \
  --temp 0.7 --top-p 0.8 --top-k 20

Option 4: Python (llama-cpp-python)

from llama_cpp import Llama

llm = Llama(
    model_path="./models/rtila-assistant-mini.Q4_K_M.gguf",
    n_ctx=2048,
    n_gpu_layers=-1,  # Use GPU/Metal if available, 0 for CPU only
)

response = llm.create_chat_completion(
    messages=[
        {"role": "system", "content": "You are RTILA Assistant..."},
        {"role": "user", "content": "Create a config for scraping product data"}
    ],
    temperature=0.7,
    top_p=0.8,
    top_k=20,
)
print(response['choices'][0]['message']['content'])

---

⚙️ Recommended Generation Settings

Based on Qwen3's official recommendations for non-thinking mode:

Parameter	Value
Temperature	0.7
Top-P	0.8
Top-K	20
Min-P	0

ℹ️ Note: This model was trained with thinking mode disabled for efficient structured JSON output. It will not produce <think>...</think> blocks.

---

💬 Example Prompts & Outputs

Example 1: Basic Extraction

Prompt:

Extract all product names and prices from the current page

Output:

{
  "settings": {
    "name": "Product Extraction"
  },
  "datasets": {
    "products": {
      "selector": ".product-item",
      "type": "list",
      "fields": {
        "name": {
          "selector": ".product-title",
          "type": "text"
        },
        "price": {
          "selector": ".product-price",
          "type": "text",
          "transform": "extractNumber"
        }
      }
    }
  },
  "commands": [
    {"action": "extract", "dataset": "products"}
  ]
}

Example 2: Pagination Loop

Prompt:

Scrape reviews from multiple pages, clicking "Next" until no more pages

Output:

{
  "settings": {
    "name": "Review Scraper with Pagination"
  },
  "datasets": {
    "reviews": {
      "selector": ".review-card",
      "type": "list",
      "fields": {
        "author": {"selector": ".reviewer-name", "type": "text"},
        "rating": {"selector": ".star-rating", "type": "attribute", "attribute": "data-rating"},
        "text": {"selector": ".review-body", "type": "text"},
        "date": {"selector": ".review-date", "type": "text"}
      }
    }
  },
  "commands": [
    {
      "action": "loop",
      "while": {"exists": ".next-page:not(.disabled)"},
      "maxIterations": 50,
      "commands": [
        {"action": "extract", "dataset": "reviews", "append": true},
        {"action": "click", "selector": ".next-page"},
        {"action": "wait", "duration": 2000}
      ]
    }
  ]
}

Example 3: Database Integration

Prompt:

After extraction, save data to PostgreSQL and send Slack notification

Output:

{
  "triggers": {
    "onComplete": [
      {
        "type": "database",
        "config": {
          "driver": "postgresql",
          "connection": "{{env.DATABASE_URL}}",
          "table": "scraped_data",
          "mode": "upsert",
          "uniqueKey": "id"
        }
      },
      {
        "type": "webhook",
        "config": {
          "url": "{{env.SLACK_WEBHOOK}}",
          "method": "POST",
          "body": {
            "text": "✅ Extraction complete: {{datasets.products.length}} items saved"
          }
        }
      }
    ]
  }
}

Example 4: Knowledge Question

Prompt:

What is the difference between execute_script and run_script?

Output:

execute_script and run_script serve different purposes in RTILA:

**execute_script**: Runs JavaScript synchronously in the page context. Returns immediately 
with the result. Best for quick DOM queries, getting values, or simple manipulations.

**run_script**: Executes JavaScript asynchronously with full control flow. Supports 
await, can make fetch requests, and handles complex async operations. Returns a Promise.

Use execute_script for: Reading values, checking conditions, simple DOM changes
Use run_script for: API calls, complex async workflows, operations that need to wait

---

🏋️ Training Details

Parameter	Value
Base Model	unsloth/Qwen3-4B
Method	QLoRA (4-bit)
LoRA Rank	128
LoRA Alpha	256
Context Length	2048 tokens
Training Examples	~400
Epochs	6 (with early stopping)
Learning Rate	2e-4
Thinking Mode	Disabled

Optimizations for Mini Version

• Highest LoRA rank (128): Maximizes learning capacity for smaller base
• More training epochs (6): Compensates for smaller model capacity
• Higher learning rate (2e-4): Better convergence for small models
• Longer context (2048): Full headroom for complex configurations
• Thinking mode disabled: Clean JSON output without <think> overhead
• Rank-stabilized LoRA (rsLoRA): More stable training dynamics

Training Data

• Navigation & Interaction patterns
• Data extraction configurations
• Logic & flow control
• Triggers & integrations
• Variables & substitution
• Advanced scripting
• Error handling
• Knowledge base Q&A

---

📝 System Prompt

For best results, use this system prompt:

You are RTILA Assistant, an expert AI for generating automation configurations for the RTILA Automation Engine.

Your capabilities:
1. Generate complete JSON configurations for web automation tasks
2. Define datasets with selectors, properties, and transformations
3. Configure navigation, extraction, loops, and conditionals
4. Set up triggers for webhooks, databases, and integrations
5. Explain RTILA concepts and best practices

When generating configurations:
- Always output valid JSON with proper structure
- Include 'settings', 'datasets', and 'commands' sections as needed
- Use appropriate selectors (CSS, XPath) for the target elements
- Apply transformations when data cleaning is required

When answering questions:
- Be concise and accurate
- Provide examples when helpful
- Reference specific RTILA features and commands

---

🔗 Model Family

Model	Link	Best For
RTILA Assistant	huggingface.co/rtila-corporation/rtila-assistant	Maximum quality
RTILA Assistant Lite	huggingface.co/rtila-corporation/rtila-assistant-lite	Mid-range devices
RTILA Assistant Mini (this)	huggingface.co/rtila-corporation/rtila-assistant-mini	Mac M1 8GB, low VRAM

RTILA Platform: rtila.com

---

📄 License

Apache 2.0

---

🙏 Acknowledgments

• Built on Qwen3-4B by Alibaba
• Fine-tuned with Unsloth
• Quantized with llama.cpp