Gemma 3 4b It Gguf

🚀 Gemma-3 4B Instruct GGUF Models

This project provides Gemma-3 4B Instruct GGUF models, which support image-text-to-text tasks. It offers experimental support for Gemma 3 Vision in llama.cpp, allowing users to perform various text generation and image understanding tasks.

🚀 Quick Start

How to Use Gemma 3 Vision with llama.cpp

To utilize the experimental support for Gemma 3 Vision in llama.cpp, follow these steps:

1. Clone the latest llama.cpp Repository:

   git clone https://github.com/ggml-org/llama.cpp.git
   cd llama.cpp
   

2. Build the Llama.cpp: Build llama.cpp as usual: Building the project. Once llama.cpp is built, copy the ./llama.cpp/build/bin/llama-gemma3-cli to a chosen folder.

3. Download the Gemma 3 gguf file: Gemma 3 gguf files. Choose a gguf file without the mmproj in the name. Example gguf file: google_gemma-3-4b-it-q4_k_l.gguf. Copy this file to your chosen folder.

4. Download the Gemma 3 mmproj file Gemma 3 mmproj files. Choose a file with mmproj in the name. Example mmproj file: google_gemma-3-4b-it-mmproj-bf16.gguf. Copy this file to your chosen folder.

5. Copy images to the same folder as the gguf files or alter paths appropriately. In the example below, the gguf files, images, and llama-gemma-cli are in the same folder. Example image: car-1.jpg. Copy this file to your chosen folder.

6. Run the CLI Tool: From your chosen folder:

   llama-gemma3-cli -m google_gemma-3-4b-it-q4_k_l.gguf --mmproj google_gemma-3-4b-it-mmproj-bf16.gguf
   

Running in chat mode, available commands:
  /image <path>    load an image
  /clear           clear the chat history
  /quit or /exit   exit the program

> /image car-1.jpg
Encoding image car-1.jpg
Image encoded in 46305 ms
Image decoded in 19302 ms

> what is the image of
Here's a breakdown of what's in the image:

**Subject:** The primary subject is a black Porsche Panamera Turbo driving on a highway.

**Details:**

*   **Car:** It's a sleek, modern Porsche Panamera Turbo, identifiable by its distinctive rear design, the "PORSCHE" lettering, and the "Panamera Turbo" badge. The license plate reads "CVC-911".
*   **Setting:** The car is on a multi-lane highway, with a blurred background of trees, a distant building, and a cloudy sky. The lighting suggests it's either dusk or dawn.
*   **Motion:** The image captures the car in motion, with a slight motion blur to convey speed.

**Overall Impression:** The image conveys a sense of speed, luxury, and power. It's a well-composed shot that highlights the car's design and performance.

Do you want me to describe any specific aspect of the image in more detail, or perhaps analyze its composition?

✨ Features

Testing and Available AI Assistants

• Testing Function Calling: Experimenting with function calling against a network monitoring service using small open source models, exploring the question "How small can it go and still function".
• Available AI Assistants:
  • 🟡 TestLLM: Runs Phi-4-mini-instruct using phi-4-mini-q4_0.gguf, llama.cpp on 6 threads of a CPU VM. It takes about 15s to load, has a slow inference speed, and processes one user prompt at a time.
  • 🟢 TurboLLM: Uses gpt-4o-mini, which is fast. Note that tokens are limited since OpenAI models are pricey. You can Login or Download the Free Network Monitor agent to get more tokens, or alternatively use the TestLLM.
  • 🔵 HugLLM: Runs open-source Hugging Face models, which are fast. It runs small models (≈8B), so the quality is lower. You can get 2x more tokens (subject to Hugging Face API availability).

Model Format Selection

• BF16 (Brain Float 16):
  • A 16-bit floating-point format for faster computation with good precision.
  • Similar dynamic range as FP32 but lower memory usage.
  • Recommended for hardware with BF16 acceleration.
  • Ideal for high-performance inference with reduced memory footprint.
• F16 (Float 16):
  • A 16-bit floating-point format with high precision but less range than BF16.
  • Works on most devices with FP16 acceleration support.
  • Slightly lower numerical precision than BF16 but generally sufficient for inference.
• Quantized Models (Q4_K, Q6_K, Q8, etc.):
  • Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
  • Lower-bit models (Q4_K) are best for minimal memory usage but may have lower precision.
  • Higher-bit models (Q6_K, Q8_0) offer better accuracy but require more memory.

Summary Table: Model Format Selection

Included Files & Details

• google_gemma-3-4b-it-bf16.gguf:
  • Model weights preserved in BF16.
  • Use for requantizing the model into a different format.
  • Best for devices with BF16 acceleration.
• google_gemma-3-4b-it-f16.gguf:
  • Model weights stored in F16.
  • Use if your device supports FP16, especially if BF16 is not available.
• google_gemma-3-4b-it-bf16-q8.gguf:
  • Output & embeddings remain in BF16.
  • All other layers quantized to Q8_0.
  • Use if your device supports BF16 and you want a quantized version.
• google_gemma-3-4b-it-f16-q8.gguf:
  • Output & embeddings remain in F16.
  • All other layers quantized to Q8_0.
• google_gemma-3-4b-it-q4_k_l.gguf:
  • Output & embeddings quantized to Q8_0.
  • All other layers quantized to Q4_K.
  • Good for CPU inference with limited memory.
• google_gemma-3-4b-it-q4_k_m.gguf:
  • Similar to Q4_K.
  • Another option for low-memory CPU inference.
• google_gemma-3-4b-it-q4_k_s.gguf:
  • Smallest Q4_K variant, using less memory at the cost of accuracy.
  • Best for very low-memory setups.
• google_gemma-3-4b-it-q6_k_l.gguf:
  • Output & embeddings quantized to Q8_0.
  • All other layers quantized to Q6_K.
• google_gemma-3-4b-it-q6_k_m.gguf:
  • A mid-range Q6_K quantized model for balanced performance.
  • Suitable for CPU-based inference with moderate memory.
• google_gemma-3-4b-it-q8.gguf:
  • Fully Q8 quantized model for better accuracy.
  • Requires more memory but offers higher precision.

📚 Documentation

Gemma 3 model card

• Model Page: Gemma
• Resources and Technical Documentation:
  • [Gemma 3 Technical Report][g3-tech-report]
  • [Responsible Generative AI Toolkit][rai-toolkit]
  • [Gemma on Kaggle][kaggle-gemma]
  • [Gemma on Vertex Model Garden][vertex-mg-gemma3]
• Terms of Use: [Terms][terms]
• Authors: Google DeepMind

Model Information

• Description: Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. Gemma 3 models are multimodal, handling text and image input and generating text output, with open weights for both pre-trained variants and instruction-tuned variants. Gemma 3 has a large, 128K context window, multilingual support in over 140 languages, and is available in more sizes than previous versions. Gemma 3 models are well-suited for a variety of text generation and image understanding tasks, including question answering, summarization, and reasoning. Their relatively small size makes it possible to deploy them in environments with limited resources such as laptops, desktops, or your own cloud infrastructure, democratizing access to state-of-the-art AI models and helping foster innovation for everyone.
• Inputs and outputs:
  • Input:
    • Text string, such as a question, a prompt, or a document to be summarized.
    • Images, normalized to 896 x 896 resolution and encoded to 256 tokens each.
    • Total input context of 128K tokens for the 4B, 12B, and 27B sizes, and 32K tokens for the 1B size.
  • Output:
    • Generated text in response to the input, such as an answer to a question, analysis of image content, or a summary of a document.
    • Total output context of 8192 tokens.

📄 License

The license for this project is Gemma.

Credits

Thanks Bartowski for the imartix upload and guidance on quantization, which has enabled the production of these gguf files.