Instructions to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ with Transformers:

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ", dtype="auto")

llama-cpp-python

How to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ",
	filename="MoQ-Quants/MoQ-3.0.gguf",
)

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

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ 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 w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16
# Run inference directly in the terminal:
llama cli -hf w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16
# Run inference directly in the terminal:
llama cli -hf w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

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 w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16
# Run inference directly in the terminal:
./llama-cli -hf w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

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 w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

Use Docker

docker model run hf.co/w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

LM Studio
Jan
Ollama
How to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ with Ollama:
```
ollama run hf.co/w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16
```

Unsloth Studio

How to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ 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 w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ 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 w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ to start chatting

How to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

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": "w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

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 w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

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 "w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16" \
  --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 w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ with Docker Model Runner:
```
docker model run hf.co/w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16
```

Lemonade

How to use w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull w-ahmad/Gemma-4-12B-it-qat-GGUF-MoQ:BF16

Run and chat with the model

lemonade run user.Gemma-4-12B-it-qat-GGUF-MoQ-BF16

List all available models

lemonade list

w-ahmad commited on Jun 9

Commit

99fd2e8

verified ·

1 Parent(s): f47fb2b

Upload MoQ-Tensor-Configs/MoQ_gemma4_12b_it_tensors_4.5.txt with huggingface_hub

Browse files

Files changed (1) hide show

MoQ-Tensor-Configs/MoQ_gemma4_12b_it_tensors_4.5.txt +329 -0

MoQ-Tensor-Configs/MoQ_gemma4_12b_it_tensors_4.5.txt ADDED Viewed

	@@ -0,0 +1,329 @@

+token_embd.weight=IQ4_NL
+blk.0.ffn_down.weight=Q4_K
+blk.0.ffn_gate.weight=IQ4_XS
+blk.0.ffn_up.weight=IQ4_XS
+blk.0.attn_k.weight=IQ4_XS
+blk.0.attn_output.weight=Q4_K
+blk.0.attn_q.weight=IQ4_XS
+blk.0.attn_v.weight=Q5_K
+blk.1.ffn_down.weight=Q4_K
+blk.1.ffn_gate.weight=IQ4_XS
+blk.1.ffn_up.weight=IQ4_XS
+blk.1.attn_k.weight=IQ4_XS
+blk.1.attn_output.weight=Q4_K
+blk.1.attn_q.weight=IQ4_XS
+blk.1.attn_v.weight=Q5_K
+blk.10.ffn_down.weight=Q4_K
+blk.10.ffn_gate.weight=IQ4_XS
+blk.10.ffn_up.weight=IQ4_XS
+blk.10.attn_k.weight=IQ4_XS
+blk.10.attn_output.weight=Q4_K
+blk.10.attn_q.weight=IQ4_XS
+blk.10.attn_v.weight=Q5_K
+blk.11.ffn_down.weight=Q4_K
+blk.11.ffn_gate.weight=IQ4_XS
+blk.11.ffn_up.weight=IQ4_XS
+blk.11.attn_k.weight=IQ4_XS
+blk.11.attn_output.weight=Q4_K
+blk.11.attn_q.weight=IQ4_XS
+blk.12.ffn_down.weight=Q4_K
+blk.12.ffn_gate.weight=IQ4_XS
+blk.12.ffn_up.weight=IQ4_XS
+blk.12.attn_k.weight=IQ4_XS
+blk.12.attn_output.weight=Q4_K
+blk.12.attn_q.weight=IQ4_XS
+blk.12.attn_v.weight=Q5_K
+blk.13.ffn_down.weight=Q4_K
+blk.13.ffn_gate.weight=IQ4_XS
+blk.13.ffn_up.weight=IQ4_XS
+blk.13.attn_k.weight=IQ4_XS
+blk.13.attn_output.weight=Q4_K
+blk.13.attn_q.weight=IQ4_XS
+blk.13.attn_v.weight=Q5_K
+blk.14.ffn_down.weight=Q4_K
+blk.14.ffn_gate.weight=IQ4_XS
+blk.14.ffn_up.weight=IQ4_XS
+blk.14.attn_k.weight=IQ4_XS
+blk.14.attn_output.weight=Q4_K
+blk.14.attn_q.weight=IQ4_XS
+blk.14.attn_v.weight=Q5_K
+blk.15.ffn_down.weight=Q4_K
+blk.15.ffn_gate.weight=IQ4_XS
+blk.15.ffn_up.weight=IQ4_XS
+blk.15.attn_k.weight=IQ4_XS
+blk.15.attn_output.weight=Q4_K
+blk.15.attn_q.weight=IQ4_XS
+blk.15.attn_v.weight=Q5_K
+blk.16.ffn_down.weight=Q4_K
+blk.16.ffn_gate.weight=IQ4_XS
+blk.16.ffn_up.weight=IQ4_XS
+blk.16.attn_k.weight=IQ4_XS
+blk.16.attn_output.weight=Q4_K
+blk.16.attn_q.weight=IQ4_XS
+blk.16.attn_v.weight=Q5_K
+blk.17.ffn_down.weight=Q4_K
+blk.17.ffn_gate.weight=IQ4_XS
+blk.17.ffn_up.weight=IQ4_XS
+blk.17.attn_k.weight=IQ4_XS
+blk.17.attn_output.weight=Q4_K
+blk.17.attn_q.weight=IQ4_XS
+blk.18.ffn_down.weight=Q4_K
+blk.18.ffn_gate.weight=IQ4_XS
+blk.18.ffn_up.weight=IQ4_XS
+blk.18.attn_k.weight=IQ4_XS
+blk.18.attn_output.weight=Q4_K
+blk.18.attn_q.weight=IQ4_XS
+blk.18.attn_v.weight=Q5_K
+blk.19.ffn_down.weight=Q4_K
+blk.19.ffn_gate.weight=IQ4_XS
+blk.19.ffn_up.weight=IQ4_XS
+blk.19.attn_k.weight=IQ4_XS
+blk.19.attn_output.weight=Q4_K
+blk.19.attn_q.weight=IQ4_XS
+blk.19.attn_v.weight=Q5_K
+blk.2.ffn_down.weight=Q4_K
+blk.2.ffn_gate.weight=IQ4_XS
+blk.2.ffn_up.weight=IQ4_XS
+blk.2.attn_k.weight=IQ4_XS
+blk.2.attn_output.weight=Q4_K
+blk.2.attn_q.weight=IQ4_XS
+blk.2.attn_v.weight=Q5_K
+blk.20.ffn_down.weight=Q4_K
+blk.20.ffn_gate.weight=IQ4_XS
+blk.20.ffn_up.weight=IQ4_XS
+blk.20.attn_k.weight=IQ4_XS
+blk.20.attn_output.weight=Q4_K
+blk.20.attn_q.weight=IQ4_XS
+blk.20.attn_v.weight=Q5_K
+blk.21.ffn_down.weight=Q4_K
+blk.21.ffn_gate.weight=IQ4_XS
+blk.21.ffn_up.weight=IQ4_XS
+blk.21.attn_k.weight=IQ4_XS
+blk.21.attn_output.weight=Q4_K
+blk.21.attn_q.weight=IQ4_XS
+blk.21.attn_v.weight=Q5_K
+blk.22.ffn_down.weight=Q4_K
+blk.22.ffn_gate.weight=IQ4_XS
+blk.22.ffn_up.weight=IQ4_XS
+blk.22.attn_k.weight=IQ4_XS
+blk.22.attn_output.weight=Q4_K
+blk.22.attn_q.weight=IQ4_XS
+blk.22.attn_v.weight=Q5_K
+blk.23.ffn_down.weight=Q4_K
+blk.23.ffn_gate.weight=IQ4_XS
+blk.23.ffn_up.weight=IQ4_XS
+blk.23.attn_k.weight=IQ4_XS
+blk.23.attn_output.weight=Q4_K
+blk.23.attn_q.weight=IQ4_XS
+blk.24.ffn_down.weight=Q4_K
+blk.24.ffn_gate.weight=IQ4_XS
+blk.24.ffn_up.weight=IQ4_XS
+blk.24.attn_k.weight=IQ4_XS
+blk.24.attn_output.weight=Q4_K
+blk.24.attn_q.weight=IQ4_XS
+blk.24.attn_v.weight=Q5_K
+blk.25.ffn_down.weight=Q4_K
+blk.25.ffn_gate.weight=IQ4_XS
+blk.25.ffn_up.weight=IQ4_XS
+blk.25.attn_k.weight=IQ4_XS
+blk.25.attn_output.weight=Q4_K
+blk.25.attn_q.weight=IQ4_XS
+blk.25.attn_v.weight=Q5_K
+blk.26.ffn_down.weight=Q4_K
+blk.26.ffn_gate.weight=IQ4_XS
+blk.26.ffn_up.weight=IQ4_XS
+blk.26.attn_k.weight=IQ4_XS
+blk.26.attn_output.weight=Q4_K
+blk.26.attn_q.weight=IQ4_XS
+blk.26.attn_v.weight=Q5_K
+blk.27.ffn_down.weight=Q4_K
+blk.27.ffn_gate.weight=IQ4_XS
+blk.27.ffn_up.weight=IQ4_XS
+blk.27.attn_k.weight=IQ4_XS
+blk.27.attn_output.weight=Q4_K
+blk.27.attn_q.weight=IQ4_XS
+blk.27.attn_v.weight=Q5_K
+blk.28.ffn_down.weight=Q4_K
+blk.28.ffn_gate.weight=IQ4_XS
+blk.28.ffn_up.weight=IQ4_XS
+blk.28.attn_k.weight=IQ4_XS
+blk.28.attn_output.weight=Q4_K
+blk.28.attn_q.weight=IQ4_XS
+blk.28.attn_v.weight=Q5_K
+blk.29.ffn_down.weight=Q4_K
+blk.29.ffn_gate.weight=IQ4_XS
+blk.29.ffn_up.weight=IQ4_XS
+blk.29.attn_k.weight=IQ4_XS
+blk.29.attn_output.weight=Q4_K
+blk.29.attn_q.weight=IQ4_XS
+blk.3.ffn_down.weight=Q4_K
+blk.3.ffn_gate.weight=IQ4_XS
+blk.3.ffn_up.weight=IQ4_XS
+blk.3.attn_k.weight=IQ4_XS
+blk.3.attn_output.weight=Q4_K
+blk.3.attn_q.weight=IQ4_XS
+blk.3.attn_v.weight=Q5_K
+blk.30.ffn_down.weight=Q4_K
+blk.30.ffn_gate.weight=IQ4_XS
+blk.30.ffn_up.weight=IQ4_XS
+blk.30.attn_k.weight=IQ4_XS
+blk.30.attn_output.weight=Q4_K
+blk.30.attn_q.weight=IQ4_XS
+blk.30.attn_v.weight=Q5_K
+blk.31.ffn_down.weight=Q4_K
+blk.31.ffn_gate.weight=IQ4_XS
+blk.31.ffn_up.weight=IQ4_XS
+blk.31.attn_k.weight=IQ4_XS
+blk.31.attn_output.weight=Q4_K
+blk.31.attn_q.weight=IQ4_XS
+blk.31.attn_v.weight=Q5_K
+blk.32.ffn_down.weight=Q4_K
+blk.32.ffn_gate.weight=IQ4_XS
+blk.32.ffn_up.weight=IQ4_XS
+blk.32.attn_k.weight=IQ4_XS
+blk.32.attn_output.weight=Q4_K
+blk.32.attn_q.weight=IQ4_XS
+blk.32.attn_v.weight=Q5_K
+blk.33.ffn_down.weight=Q4_K
+blk.33.ffn_gate.weight=IQ4_XS
+blk.33.ffn_up.weight=IQ4_XS
+blk.33.attn_k.weight=IQ4_XS
+blk.33.attn_output.weight=Q4_K
+blk.33.attn_q.weight=IQ4_XS
+blk.33.attn_v.weight=Q5_K
+blk.34.ffn_down.weight=Q4_K
+blk.34.ffn_gate.weight=IQ4_XS
+blk.34.ffn_up.weight=IQ4_XS
+blk.34.attn_k.weight=IQ4_XS
+blk.34.attn_output.weight=Q4_K
+blk.34.attn_q.weight=IQ4_XS
+blk.34.attn_v.weight=Q5_K
+blk.35.ffn_down.weight=Q4_K
+blk.35.ffn_gate.weight=IQ4_XS
+blk.35.ffn_up.weight=IQ4_XS
+blk.35.attn_k.weight=IQ4_XS
+blk.35.attn_output.weight=Q4_K
+blk.35.attn_q.weight=IQ4_XS
+blk.36.ffn_down.weight=Q4_K
+blk.36.ffn_gate.weight=IQ4_XS
+blk.36.ffn_up.weight=IQ4_XS
+blk.36.attn_k.weight=IQ4_XS
+blk.36.attn_output.weight=Q4_K
+blk.36.attn_q.weight=IQ4_XS
+blk.36.attn_v.weight=Q5_K
+blk.37.ffn_down.weight=Q4_K
+blk.37.ffn_gate.weight=IQ4_XS
+blk.37.ffn_up.weight=IQ4_XS
+blk.37.attn_k.weight=IQ4_XS
+blk.37.attn_output.weight=Q4_K
+blk.37.attn_q.weight=IQ4_XS
+blk.37.attn_v.weight=Q5_K
+blk.38.ffn_down.weight=Q4_K
+blk.38.ffn_gate.weight=IQ4_XS
+blk.38.ffn_up.weight=IQ4_XS
+blk.38.attn_k.weight=IQ4_XS
+blk.38.attn_output.weight=Q4_K
+blk.38.attn_q.weight=IQ4_XS
+blk.38.attn_v.weight=Q5_K
+blk.39.ffn_down.weight=Q4_K
+blk.39.ffn_gate.weight=IQ4_XS
+blk.39.ffn_up.weight=IQ4_XS
+blk.39.attn_k.weight=IQ4_XS
+blk.39.attn_output.weight=Q4_K
+blk.39.attn_q.weight=IQ4_XS
+blk.39.attn_v.weight=Q5_K
+blk.4.ffn_down.weight=Q4_K
+blk.4.ffn_gate.weight=IQ4_XS
+blk.4.ffn_up.weight=IQ4_XS
+blk.4.attn_k.weight=IQ4_XS
+blk.4.attn_output.weight=Q4_K
+blk.4.attn_q.weight=IQ4_XS
+blk.4.attn_v.weight=Q5_K
+blk.40.ffn_down.weight=Q4_K
+blk.40.ffn_gate.weight=IQ4_XS
+blk.40.ffn_up.weight=IQ4_XS
+blk.40.attn_k.weight=IQ4_XS
+blk.40.attn_output.weight=Q4_K
+blk.40.attn_q.weight=IQ4_XS
+blk.40.attn_v.weight=Q5_K
+blk.41.ffn_down.weight=Q4_K
+blk.41.ffn_gate.weight=IQ4_XS
+blk.41.ffn_up.weight=IQ4_XS
+blk.41.attn_k.weight=IQ4_XS
+blk.41.attn_output.weight=Q4_K
+blk.41.attn_q.weight=IQ4_XS
+blk.42.ffn_down.weight=Q4_K
+blk.42.ffn_gate.weight=IQ4_XS
+blk.42.ffn_up.weight=IQ4_XS
+blk.42.attn_k.weight=IQ4_XS
+blk.42.attn_output.weight=Q4_K
+blk.42.attn_q.weight=IQ4_XS
+blk.42.attn_v.weight=Q5_K
+blk.43.ffn_down.weight=Q4_K
+blk.43.ffn_gate.weight=IQ4_XS
+blk.43.ffn_up.weight=IQ4_XS
+blk.43.attn_k.weight=IQ4_XS
+blk.43.attn_output.weight=Q4_K
+blk.43.attn_q.weight=IQ4_XS
+blk.43.attn_v.weight=Q5_K
+blk.44.ffn_down.weight=Q4_K
+blk.44.ffn_gate.weight=IQ4_XS
+blk.44.ffn_up.weight=IQ4_XS
+blk.44.attn_k.weight=IQ4_XS
+blk.44.attn_output.weight=Q4_K
+blk.44.attn_q.weight=IQ4_XS
+blk.44.attn_v.weight=Q5_K
+blk.45.ffn_down.weight=Q4_K
+blk.45.ffn_gate.weight=IQ4_XS
+blk.45.ffn_up.weight=IQ4_XS
+blk.45.attn_k.weight=IQ4_XS
+blk.45.attn_output.weight=Q4_K
+blk.45.attn_q.weight=IQ4_XS
+blk.45.attn_v.weight=Q5_K
+blk.46.ffn_down.weight=Q4_K
+blk.46.ffn_gate.weight=IQ4_XS
+blk.46.ffn_up.weight=IQ4_XS
+blk.46.attn_k.weight=IQ4_XS
+blk.46.attn_output.weight=Q4_K
+blk.46.attn_q.weight=IQ4_XS
+blk.46.attn_v.weight=Q5_K
+blk.47.ffn_down.weight=Q4_K
+blk.47.ffn_gate.weight=IQ4_XS
+blk.47.ffn_up.weight=IQ4_XS
+blk.47.attn_k.weight=IQ4_XS
+blk.47.attn_output.weight=Q4_K
+blk.47.attn_q.weight=IQ4_XS
+blk.5.ffn_down.weight=Q4_K
+blk.5.ffn_gate.weight=IQ4_XS
+blk.5.ffn_up.weight=IQ4_XS
+blk.5.attn_k.weight=IQ4_XS
+blk.5.attn_output.weight=Q4_K
+blk.5.attn_q.weight=IQ4_XS
+blk.6.ffn_down.weight=Q4_K
+blk.6.ffn_gate.weight=IQ4_XS
+blk.6.ffn_up.weight=IQ4_XS
+blk.6.attn_k.weight=IQ4_XS
+blk.6.attn_output.weight=Q4_K
+blk.6.attn_q.weight=IQ4_XS
+blk.6.attn_v.weight=Q5_K
+blk.7.ffn_down.weight=Q4_K
+blk.7.ffn_gate.weight=IQ4_XS
+blk.7.ffn_up.weight=IQ4_XS
+blk.7.attn_k.weight=IQ4_XS
+blk.7.attn_output.weight=Q4_K
+blk.7.attn_q.weight=IQ4_XS
+blk.7.attn_v.weight=Q5_K
+blk.8.ffn_down.weight=Q4_K
+blk.8.ffn_gate.weight=IQ4_XS
+blk.8.ffn_up.weight=IQ4_XS
+blk.8.attn_k.weight=IQ4_XS
+blk.8.attn_output.weight=Q4_K
+blk.8.attn_q.weight=IQ4_XS
+blk.8.attn_v.weight=Q5_K
+blk.9.ffn_down.weight=Q4_K
+blk.9.ffn_gate.weight=IQ4_XS
+blk.9.ffn_up.weight=IQ4_XS
+blk.9.attn_k.weight=IQ4_XS
+blk.9.attn_output.weight=Q4_K
+blk.9.attn_q.weight=IQ4_XS
+blk.9.attn_v.weight=Q5_K