Instructions to use lianghsun/F1-24B-Reasoner with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use lianghsun/F1-24B-Reasoner with Transformers:

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

pipe = pipeline("text-generation", model="lianghsun/F1-24B-Reasoner")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("lianghsun/F1-24B-Reasoner")
model = AutoModelForCausalLM.from_pretrained("lianghsun/F1-24B-Reasoner")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use lianghsun/F1-24B-Reasoner with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "lianghsun/F1-24B-Reasoner"
# 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/F1-24B-Reasoner",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/lianghsun/F1-24B-Reasoner

SGLang

How to use lianghsun/F1-24B-Reasoner 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 "lianghsun/F1-24B-Reasoner" \
    --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": "lianghsun/F1-24B-Reasoner",
		"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 "lianghsun/F1-24B-Reasoner" \
        --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": "lianghsun/F1-24B-Reasoner",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use lianghsun/F1-24B-Reasoner with Docker Model Runner:
```
docker model run hf.co/lianghsun/F1-24B-Reasoner
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Model Card for F1-24B-Reasoner

F1-24B-Reasoner 是 Formosa-1（F1） 系列的 24B 推理（reasoning）模型，建構於 F1-24B-Base 之上，使用帶 <think>...</think> 思考段落之繁中 reasoning 訓練資料完成 SFT，能在回答前先輸出推理過程，提升其在多步推理、條件判斷、結構化輸出等任務上的穩定度。

⚠️ 規格重點：本模型為 24B 參數、純文本單模態，回應前段為 <think>...</think> 推理區段，後段為最終答案。

Model Details

通用的繁中對話模型在多步推理任務上常缺乏穩定的「先思考、再回答」流程，導致最終答案邏輯跳步或錯誤。F1-24B-Reasoner 透過 reasoning-style 對話資料微調，使 F1-24B 系列在 24B 級規模上具備可解釋的推理鏈能力。

核心特點 (Key Features)

24B 級 reasoning 模型：在能力與部署成本之間取得平衡，可作為企業級複雜推理應用的基底。
可解釋的推理段落：每次回答以 <think>...</think> 顯示推理過程，方便除錯與審視。
F1 家族成員：與 F1-24B-Instruct 等版本互補。

Model Description

Developed by: Liang Hsun Huang
Funded by: APMIC
Base model: lianghsun/F1-24B-Base
Model type: MistralForCausalLM (Transformers)
Language(s) (NLP): Traditional Chinese, English
License: MIT
Finetuned from model: lianghsun/F1-24B-Base

Model Sources

Repository: lianghsun/F1-24B-Reasoner

Citation

@misc{f1_24b_reasoner,
  title        = {F1-24B-Reasoner: A Traditional Chinese Reasoning-style Mistral-24B Model for Taiwan},
  author       = {Huang, Liang Hsun},
  year         = {2025},
  howpublished = {\url{https://huggingface.co/lianghsun/F1-24B-Reasoner}}
}