Instructions to use muzzy/GLM-5.2-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use muzzy/GLM-5.2-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="muzzy/GLM-5.2-GGUF",
	filename="GLM5.2-imatrix-muzzy.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use muzzy/GLM-5.2-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 muzzy/GLM-5.2-GGUF:Q2_K
# Run inference directly in the terminal:
llama cli -hf muzzy/GLM-5.2-GGUF:Q2_K

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf muzzy/GLM-5.2-GGUF:Q2_K
# Run inference directly in the terminal:
llama cli -hf muzzy/GLM-5.2-GGUF:Q2_K

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 muzzy/GLM-5.2-GGUF:Q2_K
# Run inference directly in the terminal:
./llama-cli -hf muzzy/GLM-5.2-GGUF:Q2_K

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 muzzy/GLM-5.2-GGUF:Q2_K
# Run inference directly in the terminal:
./build/bin/llama-cli -hf muzzy/GLM-5.2-GGUF:Q2_K

Use Docker

docker model run hf.co/muzzy/GLM-5.2-GGUF:Q2_K

LM Studio
Jan

vLLM

How to use muzzy/GLM-5.2-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "muzzy/GLM-5.2-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": "muzzy/GLM-5.2-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/muzzy/GLM-5.2-GGUF:Q2_K

Ollama
How to use muzzy/GLM-5.2-GGUF with Ollama:
```
ollama run hf.co/muzzy/GLM-5.2-GGUF:Q2_K
```

Unsloth Studio

How to use muzzy/GLM-5.2-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 muzzy/GLM-5.2-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 muzzy/GLM-5.2-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for muzzy/GLM-5.2-GGUF to start chatting

How to use muzzy/GLM-5.2-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf muzzy/GLM-5.2-GGUF:Q2_K

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": "muzzy/GLM-5.2-GGUF:Q2_K"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use muzzy/GLM-5.2-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 muzzy/GLM-5.2-GGUF:Q2_K

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 muzzy/GLM-5.2-GGUF:Q2_K

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use muzzy/GLM-5.2-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf muzzy/GLM-5.2-GGUF:Q2_K

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 "muzzy/GLM-5.2-GGUF:Q2_K" \
  --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"

Docker Model Runner
How to use muzzy/GLM-5.2-GGUF with Docker Model Runner:
```
docker model run hf.co/muzzy/GLM-5.2-GGUF:Q2_K
```

Lemonade

How to use muzzy/GLM-5.2-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull muzzy/GLM-5.2-GGUF:Q2_K

Run and chat with the model

lemonade run user.GLM-5.2-GGUF-Q2_K

List all available models

lemonade list

context poisoning ?? looping ?? how to avoid this

by pawarshardul - opened 1 day ago

Discussion

pawarshardul

1 day ago

Recommended primary for single- user retail |
| Zig | None (manual; arena- based) | Very low (no runtime; ~1 MB binary) | Explicit async; comptime; no hidden control flow; no borrow- checker- friction | N (T2;7: 2) | 8 | Non0;5) | Minimal;2:2),5 | Tiny3) | for this use (Tigerbeet5) | Niche— only if compile- domain modeling is worth the lack of SDK ecosystem |
| C++ | None (manual new/delete; UB risk; no safety; GC; GC; GC; no GC; no safety; no GC; safety) | Lowest (no runtime) | std::thread, lock-free libs (boost, lock-free; lock-free libs (boost),; UB risk (no safety net) | For0; no0; no5) | for3) | None5); lock-free0),; no safety,5; no9: undefined- behavior (no) | Mature | Overkill for retail; high crash risk (UB; no safety; manual memory) → violates "low crash probability" |
| Java/Kotlin | ZGC/G1 pauses (5-10µs to 10ms depending on regime; GC GC (non-deterministic, but 1-10ms in modern ZGC; 10-45µs in G1 young gen) | High (JVM = 0-2; ~100+; J0; 3, 0, 5; 0+1) | JVM,2; 100-150+ MB JVM footprint + | Threads + queues; mature; JFC (Java) (GC (GC (G1) (non-deterministic; pauses; 100 (100) (0; mature (mature); JVM (mature (mature) (JVM | Excellent (Kite Java SDK; official (kite- connect (Java)) | Reject— footprint violates "low resource" (100+MB JVM vs 10-20MB (0) (6). At0;5; JVM5;8 (5), 7) (3 (J0;0 vs Go 10-15MB) |
| Python (async (Python | None reference- counting + GIL; GIL serial hot loops; reference (GIL; GIL (G (GIL) (no (GIL (serial hot loops) (GIL (hot loops (GIL serial); (GIL serial) (hot; (G (hot (hot (hot); (GIL (GIL); (hot (hot) (G1; (G5 | 7) | H0 | G (hot6 (0 (G0) | GIL (0 | High (GIL serial hot hot loops) | Medium (object overhead; Python 3; 0; 3 (Python 3; overhead (object (object (G (overhead (G; overhead) (G (hot | asyncio single- thread event loop; (async (async (single (async (single- thread (async) (single (async) ( (single- thread (single) (single thread; single- thread (single thread (single (single- (single) ( GIL3 ( (single; (single (single (0); (single (single- thread;0 (single | Excellent0 thread; (single- thread) (single thread event loop); single- thread (event loop (single- thread event loop) (single- thread (event loop) (single- thread event (event) (event loop (event loop) (event (single thread (event (loop) (single thread (loop; (loop (event; ( (loop; event loop; event (loop (loop) (loop (event (loop; (loop; loop (event (event3 (loop); (loop (loop5 (loop2) | Excellent ( (loop ( G (loop ( (loop6 (loop (loop (single thread (single thread) (event loop (loop); (event loop; (event loop (GIL) (G (G (event loop); (GIL) (serial) (event loop serial (hot loops serial (hot loops); (GIL serial (hot) (loops (hot (hot) (hot loops (hot (GIL serial) (GIL serial (GIL (GIL) (event loop) (GIL (event loop) (event; (GIL) | Excellent6; (event0); (G (event (G (G (0 (event (event (2 (GIL) (hot (GIL) (hot) (5 (5) (hot (hot; (0; (0; G (G (G) (0; G; ( (G) ( (G; G) (0 (0 (G (G) ((G) (G (G ((G) (5; 5 (G (G; (5 (G ((G); G (G ((G; 0 ((0; 0 (1 (0; 5 ((9 (G;5 (5;5; ((; (0 (5 ( (8: 3 ((G (G ((G ( ((G; (G) (((G) (G ((G ((G ((G (((G (G (((G (((G ((G ((5 (G (((5 (G ((5) ((5) ( (5) ((5; (5 ( (((5 ( ( ( ( ((5 (((5 ( ((5) (((5 (( ( ( ( (((5) (( ( (( ((5 ((5) ((5 (((5) ( ((5 (((6 (, 5) ((0,9 (5 (((2 (5 (((5 ( (3 ( 9) ( ( (( (( (((5 (((5 ( ((((5 ((( (((5 ( ((((5 ((((5 ((( ( ((( ( ( (( (5 ((( ( ((( ((( ((( (((( ( (((( ((5 ((( (((( ((( (((( ((( (( (((( ( ((( (5 (((5 ((( ((( (( ( ((( (((( (((( 5 ((( ((( ( (((( ((( ((( (((( ((( ((((5 ((( (( (((( (((( ( (((( (5 ((( (((( (((( ((( (((( ((5 ((( ((( (((( (((5 ((((5 (((( ( ((((5 ((( ((((5 (((( (0 ( ((( (((5 0 ((5 ((((5 (((5) ( (0:0 (Python- | Excellent for9 ( ((((( ((( ((( ((( (((( (((((( (((( ((( (((( ((( 5 (((( (((( ((((( 5 (((( ((( ((( ((( (((( ((((5 (( ((( (((( (((( (( (((( ((( ((( ((((( (((( (( (((( (( (((( ((( (( (( (((( ((((((5 (( ( (((( (((( ((5 (((( ((( (((( ((((5 ((((((5 (((( (((((((((((5 ((( ((((( ((( (((( ( (((((( (((( (((((( ((( (((( (((( (( ((5 ((((5 ((((5 (((((((( (5,3 ((((5 | Excellent; ( ((((0.0 (((( ( ((( (, | Excellent4 ((((( ((( (((( (( (((( (( (((( (( (((5 ((((( ((5 (((( (((( (((( (((( ((( ((((( ((((((( (( (((( ((( (((( ((( (((( ((((((( (((( ((5 ((5((( (((( (( (((((( (( (((( ((( ((( ( ((((((( ((( (((( (( (((( (((( ((((( (((5 ((((( (((((5 (( (((5 (((( 5 ((5 ((( (((((( (((( ((5(((( ((5 (((( (( ((( (((((( (((( (( 5. (((4((((,7 ((( (5(, ((((5 ( (((9 ( ((8 | ( (((4 ( (( ((( (((( ((( (( ( (( (((( (( (( (( (((( (5 5 ((((((((( (((( (( ((( ((((((( ((((5 (( ( ((( (( (((( (((5 (( ((((( ( ( (((( ((( (((( ((( (((( ( ((((( (((( (((( ((((((( ((((((((((( ((( ((((( (((((5 (( ((( (( (((((( ( ((( (((( ((((( ( (((((( ((( ((( ((((5 (( ( ((( ((5((7 ((((1 (5 ( ((8 (5 (( ((((6 ( (0; (5: ((((9 (1 0((5 ((5 ((( (( ((( (((( (((((((( (( (( (((( ( (( (((5 ( (((( (((( (((( ( ((((((( (((5 ((( (( ((( (((( ((( ((( (( (((( ((((( (((( (( (((((((( ((( ((( (((( ((5 (((((5 ( ((((5 ((( 5 ( (( ((( (((( (( ((( (((( (( (( 5 (( (((((((( ((((((((((((((( (((( ( ((((((5 ( (((((((((((((5 ((((0( ((5 (( (((((((0 (5 ( ((((((0. . ( ((( (4: (((8.1) ( ((5 (( (((((((((5 ((( (((((5 (((( ((((((5 (((5 ((5 (( 5 ((5 ((0 (((0 (0 ((( (((( (( ((((( (((( (5 ( (( ( ( ((((0 (( (((0 ((((0 ((( ((((0 (0 ((( ((((5 (( ((( (((( ( (0 ( ((5 ((((( ((((((5 (((( (( ( (((( ((( (( (((( (( ((((0 (((( (((( (((5 ( ((0 ( ((((2 (6, (0 (. , ((( ((9 ( (, the0 (((0 (0,5 (8 (3: 0 ((( (((( (((0 (( ((( ((( ((((5 (((0 ((((((( ((((

pawarshardul

1 day ago

i am using following command on e5-2680v4 +2x rtx3090+ddr4-2400 1tb ram

./llama-server -m /lmstudio/muzzy/GLM-5.2-GGUF/GLM-5.2-smol-IQ2_KS-00001-of-00033.gguf --no-mmap -ngl 999 -cmoe -mla 1 -t 16 -amb 512 -b 1024 -ub 2048 -mla 1 -c 64000 -ctk f16 -khad -dsa -wgt 1 -cram -1 -muge -cuda graphs=1 --jinja --parallel 1 --chat-template-kwargs '{"reasoning_effort": "high"}' --spec-type mtp:n_max=4,p_min=0.90 --host 0.0.0.0 --port 1106 --override-kv llama.expert_used_count=int:2 --numa distribute

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