Instructions to use TJUNLP/FuxiTranyu-8B-DPO with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TJUNLP/FuxiTranyu-8B-DPO with Transformers:

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

pipe = pipeline("text-generation", model="TJUNLP/FuxiTranyu-8B-DPO", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("TJUNLP/FuxiTranyu-8B-DPO", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use TJUNLP/FuxiTranyu-8B-DPO with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TJUNLP/FuxiTranyu-8B-DPO"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TJUNLP/FuxiTranyu-8B-DPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/TJUNLP/FuxiTranyu-8B-DPO

SGLang

How to use TJUNLP/FuxiTranyu-8B-DPO 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 "TJUNLP/FuxiTranyu-8B-DPO" \
    --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": "TJUNLP/FuxiTranyu-8B-DPO",
		"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 "TJUNLP/FuxiTranyu-8B-DPO" \
        --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": "TJUNLP/FuxiTranyu-8B-DPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use TJUNLP/FuxiTranyu-8B-DPO with Docker Model Runner:
```
docker model run hf.co/TJUNLP/FuxiTranyu-8B-DPO
```

Model Card for FuxiTranyu-8B-DPO

Model Summary

FuxiTranyu-8B is an open-source multilingual large language model trained from scratch, with a specific focus on the multilinguality. It is trained on 600B tokens with a balanced data distribution across languages, exhibiting remarkable multilingual performance compared to previous multilingual LLMs like BLOOM-7B, PolyLM-13B.

FuxiTranyu supports 43 natural languages (Arabic, Bengali, Bulgarian, Burmese, Catalan, Chinese, Czech, Dutch, English, Filipino, Finnish, French, German, Greek, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Kazakh, Khmer, Korean, Kurdish, Kyrgyz, Lao, Malay, Persian, Polish, Portuguese, Romanian, Russian, Spanish, Swedish, Tamil, Tajik, Thai, Turkish, Turkmen, Ukrainian, Urdu, Uzbek, and Vietnamese) and cover 16 programming languages (Java, JavaScript, Python, PHP, C, C++, C#, TypeScript, Go, SQL, Rust, Ruby, Scala, Lua, Assembly, and Visual Basic).

FuxiTranyu-8B-DPO is trained using DPO with the UltraFeedback dataset, based on the FuxiTranyu-8B model.

More details on the data collection & processing, pretraining and fine-tuning of FuxiTranyu can be found in the technical report.

We also provide the instruction fine-tuned version at here.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model_path = "TJUNLP/FuxiTranyu-8B-DPO"

tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)

model = AutoModelForCausalLM.from_pretrained(model_path, device_map="auto", torch_dtype='auto', trust_remote_code=True)

messages = [{"role": "user", "content": "This is an input text:"}]
# format messages with the ChatML chat template
input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt").to(model.device)
# <|im_start|>user\nThis is an input text:<|im_end|>\n<|im_start|>assistant\n

output_ids = model.generate(input_ids, max_new_tokens=20)
response = tokenizer.decode(output_ids[0], skip_special_tokens=True)

print(response)

Citation info

@article{FuxiTranyu8B,
      title={FuxiTranyu: A Multilingual Large Language Model Trained with Balanced Data}, 
      author={Haoran Sun, Renren Jin, Shaoyang Xu, Leiyu Pan, Supryadi, Menglong Cui, Jiangcun Du, Yikun Lei, Lei Yang, Ling Shi, Juesi Xiao, Shaolin Zhu, and Deyi Xiong},
      journal={arxiv preprint arXiv:2408.06273},
      year={2024},
      url={https://arxiv.org/abs/2408.06273}
}

Downloads last month: 13

Safetensors

Model size

8B params

Tensor type

BF16

Datasets used to train TJUNLP/FuxiTranyu-8B-DPO

Collection including TJUNLP/FuxiTranyu-8B-DPO

FuxiTranyu

Collection

[EMNLP 2024 Industry Track] FuxiTranyu: A Multilingual Large Language Model Trained with Balanced Data • 3 items • Updated Mar 20, 2025

Paper for TJUNLP/FuxiTranyu-8B-DPO

FuxiTranyu: A Multilingual Large Language Model Trained with Balanced Data

Paper • 2408.06273 • Published Aug 12, 2024 • 10