ynklab/Qwen2.5-7B-d2d
ynklab/Qwen2.5-7B-d2d is a 7.6 billion parameter Qwen2.5-based language model developed by ynklab, specifically fine-tuned for multilingual document-level machine translation. It excels at translating entire documents between English and nine other languages, including German, Spanish, French, and Chinese, maintaining length consistency. This model is optimized for direct document-to-document translation tasks, leveraging a 32,768-token context length.
Loading preview...
Overview
ynklab/Qwen2.5-7B-d2d is a 7.6 billion parameter model, fine-tuned from Qwen/Qwen2.5-7B-Instruct, specifically for multilingual document-level machine translation. This model was developed as part of the research paper "Doc2FRC: Length-Consistent Document-Level Machine Translation via Fixed-Range Chunking" and is designed to perform direct document-to-document (d2d) translation.
Key Capabilities
- Document-Level Translation: Optimized for translating entire documents, ensuring length consistency.
- Multilingual Support: Supports translation between English and nine other languages in both directions:
- German
- Spanish
- French
- Italian
- Korean
- Dutch
- Portuguese
- Russian
- Chinese
- High Context Length: Utilizes a maximum sequence length of 32,768 tokens, suitable for longer documents.
Training Details
The model underwent full-parameter supervised fine-tuning on the sardinelab/DocBlocks dataset for 2 epochs, using a learning rate of 7e-6 and bfloat16 precision.
Usage
Users should refer to the Doc2FRC GitHub repository for exact inference scripts, prompting setup, and evaluation procedures. The recommended prompt format follows a ChatML-style structure, specifying source and target languages and text.
License
This model is released under the Apache License 2.0.