Instructions to use noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF",
	filename="LFM2-8B-A1B-MXFP4_MOE.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 noctrex/LFM2-8B-A1B-MXFP4_MOE-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 noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE
# Run inference directly in the terminal:
llama cli -hf noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE
# Run inference directly in the terminal:
llama cli -hf noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

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 noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE
# Run inference directly in the terminal:
./llama-cli -hf noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

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 noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE
# Run inference directly in the terminal:
./build/bin/llama-cli -hf noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

Use Docker

docker model run hf.co/noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

LM Studio
Jan

vLLM

How to use noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "noctrex/LFM2-8B-A1B-MXFP4_MOE-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": "noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

Ollama
How to use noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF with Ollama:
```
ollama run hf.co/noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE
```

Unsloth Studio

How to use noctrex/LFM2-8B-A1B-MXFP4_MOE-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 noctrex/LFM2-8B-A1B-MXFP4_MOE-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 noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF to start chatting

How to use noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

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": "noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use noctrex/LFM2-8B-A1B-MXFP4_MOE-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 noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

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 noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF with Docker Model Runner:
```
docker model run hf.co/noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE
```

Lemonade

How to use noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull noctrex/LFM2-8B-A1B-MXFP4_MOE-GGUF:MXFP4_MOE

Run and chat with the model

lemonade run user.LFM2-8B-A1B-MXFP4_MOE-GGUF-MXFP4_MOE

List all available models

lemonade list

noctrex commited on May 16

Commit

c4e935d

verified ·

1 Parent(s): 7553ad7

Add files using upload-large-folder tool

Browse files

Files changed (6) hide show

.gitattributes +3 -0
LFM2-8B-A1B-MXFP4_MOE.gguf +2 -2
LFM2-8B-A1B-MXFP4_MOE_BF16.gguf +3 -0
LFM2-8B-A1B-MXFP4_MOE_F16.gguf +3 -0
LFM2-8B-A1B-Q8_XL_MOE.gguf +3 -0
README.md +24 -2

.gitattributes CHANGED Viewed

@@ -34,3 +34,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 LFM2-8B-A1B-MXFP4_MOE.gguf filter=lfs diff=lfs merge=lfs -text

 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 LFM2-8B-A1B-MXFP4_MOE.gguf filter=lfs diff=lfs merge=lfs -text
+LFM2-8B-A1B-MXFP4_MOE_F16.gguf filter=lfs diff=lfs merge=lfs -text
+LFM2-8B-A1B-MXFP4_MOE_BF16.gguf filter=lfs diff=lfs merge=lfs -text
+LFM2-8B-A1B-Q8_XL_MOE.gguf filter=lfs diff=lfs merge=lfs -text

LFM2-8B-A1B-MXFP4_MOE.gguf CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6c433c7853f73bf948206a8902103e212044710c8a3c5199b11ddf022abffb12
-size 4750671040

 version https://git-lfs.github.com/spec/v1
+oid sha256:976dba5cd53212f713a5147899814ef5ec14216b65d4cf6009973111559e0cc8
+size 4876500032

LFM2-8B-A1B-MXFP4_MOE_BF16.gguf ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:e7906153f0dc2480ddbb0f3a76235713d748656148ba17f54208391459821d8e
+size 5301173312

LFM2-8B-A1B-MXFP4_MOE_F16.gguf ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:3c73f2907d0471f91214771cc80e9c62abfaf53288416ed94785c62877dc9314
+size 5301173312

LFM2-8B-A1B-Q8_XL_MOE.gguf ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:cf2f9ddad8112216d6dff9f806f0382acdc01c0f1521f4ed9b3c55ee398f9e3f
+size 9418931264

README.md CHANGED Viewed

@@ -3,6 +3,28 @@ pipeline_tag: text-generation
 base_model:
 - LiquidAI/LFM2-8B-A1B
 ---
-This is a MXFP4_MOE quantization of the model LFM2-8B-A1B
-Original model: https://huggingface.co/unsloth/LFM2-8B-A1B

 base_model:
 - LiquidAI/LFM2-8B-A1B
 ---
+These are **MXFP4** quantizations of the model [LiquidAI / LFM2-8B-A1B](https://huggingface.co/LiquidAI/LFM2-8B-A1B)
+## Quick Start
+1. Download the latest release of **llama.cpp**.
+2. Download your preferred model variant from below.
+## Which version should I choose?
+All FP4 variants use **MXFP4** for the MoE (Mixture of Experts) weights to keep the model efficient.
+I've included also a new type Q8_XL_MOE, that uses Q8 for MoE tensors and BF16 for everything else.
+The difference lies in how the remaining tensors are handled:
+| Variant | Quality | Performance | Size | Recommendation |
+| :--- | :--- | :--- | ---: | :--- |
+| **Q8_XL_MOE** | ⭐⭐⭐⭐⭐ | Variable* | 8.77GiB | Maximum quality, uses Q8 instead of FP4 for the MoE weights. |
+| **BF16** | ⭐⭐⭐ | Variable* | 4.54GiB | Best for maximum accuracy; original unquantized weights. |
+| **F16** | ⭐⭐ | Fast | 4.94GiB | Great alternative if BF16 is slow on your hardware. |
+| **Q8** | ⭐ | Fastest | 4.94GiB | Balanced performance and memory usage. |
+**Note:** On some older architectures, BF16 may be slower than F16.
+Check that your GPU supports native BF16
+Recommended parameters from LiquidAI:
+- temperature 0.3
+- min_p 0.15
+- repetition_penalty 1.05