Instructions to use ubergarm/GLM-4.7-Flash-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use ubergarm/GLM-4.7-Flash-GGUF with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ubergarm/GLM-4.7-Flash-GGUF:MXFP4
# Run inference directly in the terminal:
llama cli -hf ubergarm/GLM-4.7-Flash-GGUF:MXFP4

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ubergarm/GLM-4.7-Flash-GGUF:MXFP4
# Run inference directly in the terminal:
llama cli -hf ubergarm/GLM-4.7-Flash-GGUF:MXFP4

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 ubergarm/GLM-4.7-Flash-GGUF:MXFP4
# Run inference directly in the terminal:
./llama-cli -hf ubergarm/GLM-4.7-Flash-GGUF:MXFP4

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 ubergarm/GLM-4.7-Flash-GGUF:MXFP4
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ubergarm/GLM-4.7-Flash-GGUF:MXFP4

Use Docker

docker model run hf.co/ubergarm/GLM-4.7-Flash-GGUF:MXFP4

LM Studio
Jan

vLLM

How to use ubergarm/GLM-4.7-Flash-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ubergarm/GLM-4.7-Flash-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ubergarm/GLM-4.7-Flash-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ubergarm/GLM-4.7-Flash-GGUF:MXFP4

Ollama
How to use ubergarm/GLM-4.7-Flash-GGUF with Ollama:
```
ollama run hf.co/ubergarm/GLM-4.7-Flash-GGUF:MXFP4
```

Unsloth Studio

How to use ubergarm/GLM-4.7-Flash-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 ubergarm/GLM-4.7-Flash-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 ubergarm/GLM-4.7-Flash-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for ubergarm/GLM-4.7-Flash-GGUF to start chatting

How to use ubergarm/GLM-4.7-Flash-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ubergarm/GLM-4.7-Flash-GGUF:MXFP4

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "ubergarm/GLM-4.7-Flash-GGUF:MXFP4"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use ubergarm/GLM-4.7-Flash-GGUF with Docker Model Runner:
```
docker model run hf.co/ubergarm/GLM-4.7-Flash-GGUF:MXFP4
```

Lemonade

How to use ubergarm/GLM-4.7-Flash-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ubergarm/GLM-4.7-Flash-GGUF:MXFP4

Run and chat with the model

lemonade run user.GLM-4.7-Flash-GGUF-MXFP4

List all available models

lemonade list

Hermes Agent

How to use ubergarm/GLM-4.7-Flash-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ubergarm/GLM-4.7-Flash-GGUF:MXFP4

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default ubergarm/GLM-4.7-Flash-GGUF:MXFP4

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use ubergarm/GLM-4.7-Flash-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ubergarm/GLM-4.7-Flash-GGUF:MXFP4

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "ubergarm/GLM-4.7-Flash-GGUF:MXFP4" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Explore more quants of the REAP variant

by freenode - opened Feb 9

Discussion

freenode

Feb 9

Quants of the REAP variant of this model work very well with opencode tasks and gives me a lot more context in my 32GB VRAM: https://huggingface.co/unsloth/GLM-4.7-Flash-REAP-23B-A3B-GGUF/blob/main/GLM-4.7-Flash-REAP-23B-A3B-UD-Q6_K_XL.gguf

I would like to see if an IQ5_K of GLM-4.7-Flash-REAP could outperform Unsloth's UD-Q6_K_XL, or at least buy me more context with a minimal impact on performance. I'm working on getting AVX512 enabled on the host I have access to, so that I could try doing this myself with the immensely helpful logs/guidance that you and ik_llama have made available, but as of right now I can't. What do you think?

ubergarm

Owner Feb 9

What do you think?

Follow your heart as intrinsic motivation is the guiding factor in all activity.

Personally, I tend to use a smaller quant of the original model over a REAP, but I've not personally done head-head comparison between the two methods.

If you decide to try the experiment, go for it and please report back as I am curious what you'd find.

Check out this video for a quick overview of the quantization steps: https://blog.aifoundry.org/p/adventures-in-model-quantization
Check this old guide for some explicit old commands: https://github.com/ikawrakow/ik_llama.cpp/discussions/434
i have some logs/ folders in my newer model releases with some more updated command examples
I'd suggest if this is your first quant to start off using someone else's bf16 GGUF and imatrix file for the model you want to quantize. then adapting recipes from my model cards as desired and cooking your own quant won't take too much resources other than disk space to hold it all.
finally you can compare perplexity / kld between your custom quant and whatever other quant you want. keep in mind ppl/kld are best for relative comparisons of various quants of the same model. can't really compare numbers across different models.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment