Instructions to use Munchit/dam-vision-v3-gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Munchit/dam-vision-v3-gguf with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Munchit/dam-vision-v3-gguf",
	filename="dam_vision_mmproj_f16.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use Munchit/dam-vision-v3-gguf with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Munchit/dam-vision-v3-gguf:F16
# Run inference directly in the terminal:
llama-cli -hf Munchit/dam-vision-v3-gguf:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Munchit/dam-vision-v3-gguf:F16
# Run inference directly in the terminal:
llama-cli -hf Munchit/dam-vision-v3-gguf:F16

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 Munchit/dam-vision-v3-gguf:F16
# Run inference directly in the terminal:
./llama-cli -hf Munchit/dam-vision-v3-gguf:F16

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 Munchit/dam-vision-v3-gguf:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Munchit/dam-vision-v3-gguf:F16

Use Docker

docker model run hf.co/Munchit/dam-vision-v3-gguf:F16

LM Studio
Jan
Ollama
How to use Munchit/dam-vision-v3-gguf with Ollama:
```
ollama run hf.co/Munchit/dam-vision-v3-gguf:F16
```

Unsloth Studio new

How to use Munchit/dam-vision-v3-gguf 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 Munchit/dam-vision-v3-gguf 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 Munchit/dam-vision-v3-gguf to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Munchit/dam-vision-v3-gguf to start chatting

Docker Model Runner
How to use Munchit/dam-vision-v3-gguf with Docker Model Runner:
```
docker model run hf.co/Munchit/dam-vision-v3-gguf:F16
```

Lemonade

How to use Munchit/dam-vision-v3-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Munchit/dam-vision-v3-gguf:F16

Run and chat with the model

lemonade run user.dam-vision-v3-gguf-F16

List all available models

lemonade list

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

DAM Vision v3 — GGUF

Fine-tuned Qwen2-VL-2B-Instruct for structured video-frame analysis. Emits a 21-field JSON per frame (people, camera motion, setting, faces, etc.) for footage search inside a Digital Asset Management pipeline.

Training details

Setting	Value
Base	Qwen/Qwen2-VL-2B-Instruct
Method	QLoRA (4-bit), LLaMA-Factory
Rank / Alpha	32 / 64
LoRA targets	q_proj, k_proj, v_proj, o_proj
Image resolution	384 px
Epochs	3 + best-model
Eval loss	~2.16 (step 75 → stable)

Files

File	Description
`dam_vision_q4_k_m.gguf`	Q4_K_M quantised weights (~941 MB)
`dam_vision_mmproj_f16.gguf`	Multimodal projector, F16 (~1.3 GB)

Usage with llama.cpp / ollama

# llama.cpp (llava-cli)
llama-llava-cli \
  -m dam_vision_q4_k_m.gguf \
  --mmproj dam_vision_mmproj_f16.gguf \
  --image frame.jpg \
  -p "Analyse this frame and return JSON."

# ollama — place both files in the same dir alongside a Modelfile
ollama create dam-vision -f Modelfile
ollama run dam-vision

Downloads last month: 192

GGUF

Model size

2B params

Architecture

qwen2vl

Hardware compatibility

4-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Munchit/dam-vision-v3-gguf

Base model

Qwen/Qwen2-VL-2B

Finetuned

Qwen/Qwen2-VL-2B-Instruct

Quantized

(53)

this model