ynklab/Qwen2.5-7B-Sep_0c1t

TEXT GENERATIONPricing:Input $0.4 / Cached $0.08 / Output $0.8Concurrent Unit Cost:1Model Size:7.6BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Sep 7, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

ynklab/Qwen2.5-7B-Sep_0c1t is a 7.6 billion parameter model developed by ynklab, fine-tuned from Qwen/Qwen2.5-7B-Instruct for multilingual chunk-level machine translation. Optimized for fixed-range chunking, it supports translation between English and nine other languages (German, Spanish, French, Italian, Korean, Dutch, Portuguese, Russian, Chinese) in both directions. This model is specifically designed for document-level translation by processing individual chunks without preceding context, making it suitable for applications requiring length-consistent translation.

Loading preview...

Model Overview

ynklab/Qwen2.5-7B-Sep_0c1t is a 7.6 billion parameter language model, fine-tuned from Qwen/Qwen2.5-7B-Instruct. It is specifically designed for multilingual chunk-level machine translation as part of the Doc2FRC project, focusing on length-consistent document translation via fixed-range chunking.

Key Capabilities

  • Multilingual Translation: Supports translation between English and nine other languages (German, Spanish, French, Italian, Korean, Dutch, Portuguese, Russian, Chinese) in both directions.
  • Chunk-Level Processing: Optimized for translating individual text chunks (256–512 tokens) without relying on preceding context chunks, designated as the Sep_0c1t variant.
  • Document-Level Translation: Intended for use in a larger pipeline (Doc2FRC) where documents are first chunked, translated, and then reassembled, ensuring length consistency.
  • Fine-tuned Performance: Underwent full-parameter supervised fine-tuning for 2 epochs on the sardinelab/DocBlocks dataset, utilizing a maximum sequence length of 32,768 tokens.

When to Use This Model

This model is ideal for developers and researchers working on:

  • Document-level machine translation where maintaining chunk integrity and length consistency is crucial.
  • Multilingual applications requiring translation between English and the specified nine languages.
  • Research into fixed-range chunking and its application in machine translation, particularly when integrating with the Doc2FRC GitHub repository for full document processing workflows.

It is important to note that for full document translation, the model should be used in conjunction with the chunking and reconstruction procedures detailed in the Doc2FRC project.

Finetunes
1 models