Seed-X-Instruct-7B

Introduction

We are excited to introduce Seed-X, a powerful series of open-source multilingual translation language models, including an instruction model, a reinforcement learning model, and a reward model. It pushes the boundaries of translation capabilities within 7 billion parameters. We develop Seed-X as an accessible, off-the-shelf tool to support the community in advancing translation research and applications:

• Exceptional translation capabilities: Seed-X exhibits state-of-the-art translation capabilities, on par with or outperforming ultra-large models like Gemini-2.5, Claude-3.5, and GPT-4, as validated by human evaluations and automatic metrics.
• Deployment and inference-friendly: With a compact 7B parameter count and mistral architecture, Seed-X offers outstanding translation performance in a lightweight and efficient package, ideal for deployment and inference.
• Broad domain coverage: Seed-X excels on a highly challenging translation test set spanning diverse domains, including the internet, science and technology, office dialogues, e-commerce, biomedicine, finance, law, literature, and entertainment.

This repo contains the Seed-X-Instruct model, with the following features:

• Type: Causal language models
• Training Stage: Pretraining & Post-training
• Support: Multilingual translation among 28 languages

Languages	Abbr.	Languages	Abbr.	Languages	Abbr.	Languages	Abbr.
Arabic	ar	French	fr	Malay	ms	Russian	ru
Czech	cs	Croatian	hr	Norwegian Bokmal	nb	Swedish	sv
Danish	da	Hungarian	hu	Dutch	nl	Thai	th
German	de	Indonesian	id	Norwegian	no	Turkish	tr
English	en	Italian	it	Polish	pl	Ukrainian	uk
Spanish	es	Japanese	ja	Portuguese	pt	Vietnamese	vi
Finnish	fi	Korean	ko	Romanian	ro	Chinese	zh

Model Downloads

Model Name	Description	Download
👉 Seed-X-Instruct	Instruction-tuned for alignment with user intent.	🤗 Model
Seed-X-PPO	RL trained to boost translation capabilities.	🤗 Model
Seed-X-RM	Reward model to evaluate the quality of translation.	🤗 Model

Quickstart

Here is a simple example demonstrating how to load the model and perform translation using vllm

from vllm import LLM, SamplingParams

model_path = "./ByteDance-Seed/Seed-X-Instruct-7B"

model = LLM(model=model_path,
            max_num_seqs=512,
            tensor_parallel_size=8,
            enable_prefix_caching=True, 
            gpu_memory_utilization=0.95)

messages = [
    "Translate the following English sentence into Chinese:\nMay the force be with you <zh>", # without CoT
    "Translate the following English sentence into Chinese and explain it in detail:\nMay the force be with you <zh>" # with CoT
]

# Sampling
decoding_params = SamplingParams(temperature=0,
                                 max_tokens=512,
                                 skip_special_tokens=True)
# Beam Search
decoding_params = BeamSearchParams(beam_width=4, 
                                    max_tokens=512)

results = model.generate(messages, decoding_params)
responses = [res.outputs[0].text.strip() for res in results]

print(responses)

Evaluation

We evaluated Seed-X on a diverse set of translation benchmarks, including FLORES-200, WMT-25, and a publicly released challenge set accompanied by human evaluations. For detailed benchmark results and analysis, please refer to our Technical Report.

License

This project is licensed under OpenMDW. See the LICENSE file for details.

Citation

We will soon publish our technical report on Arxiv.