Tower-7B-Sep_1c1t

This model is released as part of our paper Doc2FRC: Length-Consistent Document-Level Machine Translation via Fixed-Range Chunking. The code and paper-specific inference scripts are available in the Doc2FRC GitHub repository.

Tower-7B-Sep_1c1t is a full-parameter fine-tuned version of Unbabel/TowerInstruct-Mistral-7B-v0.2 for multilingual chunk-level machine translation. It is the Sep_1c1t variant: each call uses one preceding source-language context chunk and translates the current source chunk into one target chunk. The training examples were derived from sardinelab/DocBlocks using fixed-range chunking.

Supported translation directions

The model supports translation between English and the following languages in both directions:

• German
• Spanish
• French
• Italian
• Korean
• Dutch
• Portuguese
• Russian
• Chinese

General usage

The example below demonstrates general model usage for translating a current source chunk with one preceding source-language context chunk. For the exact document chunking, inference scripts, prompting setup, and evaluation procedure used in the paper, please refer to the Doc2FRC GitHub repository.

Recommended prompt format

The model was fine-tuned with the following raw ChatML-style translation prompt:

<|im_start|>user
Context
{SOURCE_LANGUAGE}: {CONTEXT_TEXT}
Translate the following source text from {SOURCE_LANGUAGE} into {TARGET_LANGUAGE}.
{SOURCE_LANGUAGE}: {SOURCE_TEXT}.
{TARGET_LANGUAGE}: <|im_end|>
<|im_start|>assistant

Use full English language names such as English, Chinese, German, or Russian. The model was trained using fixed-range chunks of 256–512 tokens. CONTEXT_TEXT should contain the immediately preceding source-language chunk.

Transformers example

Install a PyTorch build appropriate for your hardware, together with Transformers and Accelerate. PyTorch 2.6 or later is recommended for loading the current PyTorch .bin checkpoint files.

pip install "transformers>=4.56.2" accelerate

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

MODEL_ID = "ynklab/Tower-7B-Sep_1c1t"

tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
model = AutoModelForCausalLM.from_pretrained(
    MODEL_ID,
    dtype=torch.bfloat16,
    device_map="auto",
)
model.eval()

source_language = "English"
target_language = "Chinese"
context_text = "We planned a picnic for this afternoon."
source_text = "The weather is nice today"

prompt = (
    "<|im_start|>user\n"
    "Context\n"
    f"{source_language}: {context_text}\n"
    f"Translate the following source text from {source_language} "
    f"into {target_language}.\n"
    f"{source_language}: {source_text}.\n"
    f"{target_language}: <|im_end|>\n"
    "<|im_start|>assistant\n"
)

# The Tower tokenizer adds the beginning-of-sequence token used during training.
inputs = tokenizer(prompt, return_tensors="pt")
inputs = {name: tensor.to(model.device) for name, tensor in inputs.items()}

with torch.inference_mode():
    outputs = model.generate(
        **inputs,
        max_new_tokens=1024,
        do_sample=False,
        repetition_penalty=1.05,
    )

generated_tokens = outputs[0, inputs["input_ids"].shape[1]:]
translation = tokenizer.decode(
    generated_tokens,
    skip_special_tokens=True,
).strip()

print(translation)

For paper-level document translation, split the document into fixed-range chunks and reconstruct the translated chunks using the procedures provided in the Doc2FRC repository.

Training

• Base model: TowerInstruct-Mistral-7B-v0.2
• Training method: full-parameter supervised fine-tuning
• Training variant: Sep_1c1t (one preceding source-language context chunk and one target chunk)
• Chunk range: 256–512 tokens
• Epochs: 2
• Learning rate: 7e-6
• Learning-rate scheduler: cosine
• Warmup steps: 125
• Maximum sequence length: 32,768 tokens
• Training precision: bfloat16
• Optimizer: AdamW
• Weight decay: 0.01

License

This model preserves the CC BY-NC-SA 4.0 License distributed with its base model, TowerInstruct-Mistral-7B-v0.2. See the LICENSE file and the upstream model card for the applicable terms.

DocBlocks contains material derived from multiple sources. Users should also consult the DocBlocks dataset and the original data sources for their applicable licensing conditions.

Acknowledgements

This model is based on TowerInstruct-Mistral-7B-v0.2 and was fine-tuned using DocBlocks. Please cite our paper when using this model in academic work.

Citation

@misc{wang2026doc2frclengthconsistentdocumentlevelmachine,
      title={Doc2FRC: Length-Consistent Document-Level Machine Translation via Fixed-Range Chunking}, 
      author={Xiaotian Wang and Youyuan Lin and Zhan Shen and Hitomi Yanaka},
      year={2026},
      eprint={2609.12674},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2609.12674}, 
}