Instructions to use lianghsun/Llama-3.2-Taiwan-3B-Instruct-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 lianghsun/Llama-3.2-Taiwan-3B-Instruct-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 lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M

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 lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M

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 lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M

Use Docker

docker model run hf.co/lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "lianghsun/Llama-3.2-Taiwan-3B-Instruct-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": "lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M

Ollama
How to use lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF with Ollama:
```
ollama run hf.co/lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M
```

Unsloth Studio

How to use lianghsun/Llama-3.2-Taiwan-3B-Instruct-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 lianghsun/Llama-3.2-Taiwan-3B-Instruct-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 lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF to start chatting

Docker Model Runner
How to use lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF with Docker Model Runner:
```
docker model run hf.co/lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M
```

Lemonade

How to use lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Llama-3.2-Taiwan-3B-Instruct-GGUF-Q4_K_M

List all available models

lemonade list

Atomic Chat

A newer version of this model is available: lianghsun/Llama-3.2-Taiwan-3B-Instruct

Model Card for Llama-3.2-Taiwan-3B-Instruct-GGUF

Llama-3.2-Taiwan-3B-Instruct-GGUF 是 Llama-3.2-Taiwan-3B-Instruct 透過 llama.cpp 轉換之 .gguf 量化版本，提供多種量化等級權重，可在 llama.cpp、Ollama、LM Studio 等推論工具中部署，適合本機端使用。

⚠️ 規格重點：本模型為 GGUF 量化版本，非原始 fp16／bf16 權重。已知問題：量化後模型有機率輸出全部簡體中文。

Model Details

繁中模型在端側部署時常受限於 GPU 記憶體；本模型對 Llama-3.2-Taiwan-3B-Instruct 進行 GGUF 量化，提供多種精度等級的權重檔，方便使用者依硬體限制與品質需求選擇對應版本。

請依不同 tag 選擇對映之原始非量化版本，最新 main 分支對映 v2025.01.01；原始（非量化）模型介紹請參考 lianghsun/Llama-3.2-Taiwan-3B-Instruct。

Model Change Log

Update Date	Model Version	Key Changes
2025-01-01	v2025.01.01	This version corresponds to the `v2025.01.01` release of Llama-3.2-Taiwan-3B-Instruct.
2024-12-11	v2024.12.11	This version corresponds to the `v2024.11.27` release of Llama-3.2-Taiwan-3B-Instruct.

核心特點 (Key Features)

多量化等級：提供 Q4、Q5、Q6、Q8 等不同精度權重，方便依硬體選擇。
本機可部署：適用於 llama.cpp、Ollama、LM Studio 等推論工具，可在筆電 CPU／Apple Silicon 流暢執行。
與母模型同步更新：每次 Llama-3.2-Taiwan-3B-Instruct 釋出新版本時同步重新量化。

Model Description

Developed by: Liang Hsun Huang
Funded by: APMIC
Base model: lianghsun/Llama-3.2-Taiwan-3B-Instruct
Model type: LlamaForCausalLM (GGUF)
Language(s) (NLP): Traditional Chinese, English（含其他多語）
License: llama3.2

Model Sources

Repository: lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF
Original Model: lianghsun/Llama-3.2-Taiwan-3B-Instruct

Known Issues

How to use in Ollama

直接以 ollama run 載入 GGUF 時可能出現「文不對題」現象，根因為預設 chat template 不正確。本 repo 內含 template 檔可解決此問題。如需自訂對話模板，請參考 Ollama Modelfile 文件。

Citation

@misc{llama_3_2_taiwan_3b_instruct_gguf,
  title        = {Llama-3.2-Taiwan-3B-Instruct-GGUF: Quantized GGUF Version of Llama-3.2-Taiwan-3B-Instruct},
  author       = {Huang, Liang Hsun},
  year         = {2024},
  howpublished = {\url{https://huggingface.co/lianghsun/Llama-3.2-Taiwan-3B-Instruct-GGUF}}
}