Instructions to use deepgrove/maple-preview-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use deepgrove/maple-preview-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="deepgrove/maple-preview-GGUF")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("deepgrove/maple-preview-GGUF", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use deepgrove/maple-preview-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 deepgrove/maple-preview-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf deepgrove/maple-preview-GGUF:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf deepgrove/maple-preview-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf deepgrove/maple-preview-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 deepgrove/maple-preview-GGUF:F16
# Run inference directly in the terminal:
./llama-cli -hf deepgrove/maple-preview-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 deepgrove/maple-preview-GGUF:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf deepgrove/maple-preview-GGUF:F16

Use Docker

docker model run hf.co/deepgrove/maple-preview-GGUF:F16

LM Studio
Jan

vLLM

How to use deepgrove/maple-preview-GGUF with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/deepgrove/maple-preview-GGUF:F16

SGLang

How to use deepgrove/maple-preview-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "deepgrove/maple-preview-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "deepgrove/maple-preview-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "deepgrove/maple-preview-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "deepgrove/maple-preview-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use deepgrove/maple-preview-GGUF with Ollama:
```
ollama run hf.co/deepgrove/maple-preview-GGUF:F16
```

Unsloth Studio

How to use deepgrove/maple-preview-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 deepgrove/maple-preview-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 deepgrove/maple-preview-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

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

How to use deepgrove/maple-preview-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf deepgrove/maple-preview-GGUF:F16

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": "deepgrove/maple-preview-GGUF:F16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use deepgrove/maple-preview-GGUF with Docker Model Runner:
```
docker model run hf.co/deepgrove/maple-preview-GGUF:F16
```

Lemonade

How to use deepgrove/maple-preview-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull deepgrove/maple-preview-GGUF:F16

Run and chat with the model

lemonade run user.maple-preview-GGUF-F16

List all available models

lemonade list

Hermes Agent

How to use deepgrove/maple-preview-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 deepgrove/maple-preview-GGUF:F16

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 deepgrove/maple-preview-GGUF:F16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use deepgrove/maple-preview-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf deepgrove/maple-preview-GGUF:F16

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 "deepgrove/maple-preview-GGUF:F16" \
  --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"

Available quantisations are confusing

by jpeg729 - opened 10 days ago

Discussion

jpeg729

10 days ago

HuggingFace lists four available quants

1-bit = 4.98Gb
2-bit = 5.9Gb
16-bit = 5.43Gb and/or 6.35Gb
How the hell is a 16 bit quant smaller than a 2-bit quant?

If you look at the file listings you see

maple-preview-TQ1_0-head-F16.gguf = 5.43Gb
maple-preview-TQ1_0-head-Q4_K.gguf = 4.89Gb
maple-preview-TQ2_0-head-F16.gguf = 6.53Gb
maple-preview-TQ2_0-head-Q4_K.gguf = 5.9Gb
Why does a TQ2 masquerade as F16 and Q4_K?

When I use the huggingface "Use this model option" it proposes llama cli -hf deepgrove/maple-preview-GGUF:F16
If I switch to TQ1_0 it downloads the maple-preview-TQ1_0-head-F16.gguf file.
If I manually type llama cli -hf deepgrove/maple-preview-GGUF:TQ1_0-head-Q4_K I can make it download the Q4_K file which is smaller, for no apparent reason.

Can someone explain the differences?

deepgrove-team

deepgrove org 8 days ago

The confusion comes from the precision of the lm-head.

Decoder layers in our model are in ternary precision - which can be represented with tq1 and tq2 (different packing methods trading off memory for speed).

The separate precision of either q4 or f16 is the precision that the lm-head runs on.

Please refer to https://github.com/deepgrove-ai/llama.cpp for more details.

deepgrove-team changed discussion status to closed 8 days ago

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