Qwen3.5-0.8B-weight-space-programming-GGUF by NotCherry

Qwen3.5-0.8B-weight-space-programming-GGUF by NotCherry

Packages the Qwen3.5-0.8B weight-space-programming model as GGUF files for llama.cpp, with a 4-bit text model and a separate vision projector for image input.

Made by NotCherry on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: This repository holds a GGUF conversion of the Qwen3.5-0.8B weight-space-programming model by RikiyaT. It is a format change only, not a new model and not changed weights. It has two files, a 4-bit text model and a vision projector for image input.

Who it suits: It suits people who run models locally with llama.cpp and want a small model that can also take images. The original model's card holds the details of what it is for.

Industry | Who uses it | How they use it | Example
Local AI | llama.cpp users | Run the quantized text model on a modest computer | Loading the Q4_K_M file in llama.cpp
Software | Developers testing image prompts | Pair the text file with the vision projector | Sending an image prompt with the matching projector file
Research | People following the original model | Try the original checkpoint in a GGUF runtime | Checking how the weights behave in llama.cpp

People who want the original safetensors weights or evaluation results should use the original model instead.

Where to use it:

• llama.cpp on Windows, macOS or Linux
• Any other tool that reads GGUF files
• Python, only for downloading the files with the Hugging Face tools

Use case | Who | What they get back
Text prompts | Local model users | Replies from the 4-bit text model
Image prompts | Developers | Replies that take an image, using the vision projector file

How to set it up:

• Python, for the download command
• llama.cpp, to run the GGUF files
• Disk space for about 386 MB of 4-bit weights, plus the projector file

1. Install the Hugging Face hub package.
2. Download the files in this repository.
3. Point llama.cpp at the text file, and add the projector file for images.

pip install -U huggingface_hub

hf download NotCherry/Qwen3.5-0.8B-weight-space-programming-GGUF

How to use it: Run the text file in llama.cpp for text prompts. For image prompts, pass the matching projector file with the --mmproj option. The card gives no full run command, so check the llama.cpp help for the rest.

Tool | What it does
Q4_K_M file | Holds the 4-bit text model, made with llama.cpp
mmproj file | Holds the F16 vision projector for image input

Example: You download the repository, load the Q4_K_M file in llama.cpp and add the projector file with --mmproj, then send an image together with a question. The card shows no sample prompt or output.

For newcomers and experts: GGUF is a file format that tools like llama.cpp can load. A first small step is to download the files and open the text file in a GGUF-capable app. For what the model was built to do, read the original card.

The conversion started from the original safetensors checkpoint, then used llama-quantize with Q4_K_M. The vision projector stays in F16 to support image input. The author makes no evaluation or quality claims for the converted files.

Original safetensors checkpoint
Converted to GGUF with llama.cpp
Text model quantized to Q4_K_M
Vision projector kept in F16

Good to know:

• License: Apache 2.0, inherited from the original model.
• Cost to run: none named. The steps need no API key or paid service.
• Data it reaches: it downloads files from the Hugging Face Hub, then runs locally in your own GGUF tool.
• Not for: judging model quality. The card makes no evaluation claims for this conversion.
• The record lists 773M parameters.

On Hugging Face, 13 November 2026

• Task: image text to text
• Library: gguf
• Parameters: 773M
• Weights: about 386 MB at 4-bit (estimate)
• Formats: GGUF
• Downloads in the last 30 days: 1,731
• Likes: 3
• License: apache-2.0
• Based on: RikiyaT/Qwen3.5-0.8B-weight-space-programming
• Tags: gguf, qwen3.5, q4_k_m, image-text-to-text, conversational

Links

• Model: https://huggingface.co/NotCherry/Qwen3.5-0.8B-weight-space-programming-GGUF
• Files: https://huggingface.co/NotCherry/Qwen3.5-0.8B-weight-space-programming-GGUF/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download NotCherry/Qwen3.5-0.8B-weight-space-programming-GGUF

Then follow the usage in its model card. Nothing in this kit runs by itself.