moutons/Llama-3.1-Swallow-JP-EN-Translator-v1-8B-mlx-fp16

TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:Apr 17, 2026License:llama3.3Architecture:Transformer Cold

The moutons/Llama-3.1-Swallow-JP-EN-Translator-v1-8B-mlx-fp16 is an 8 billion parameter model, converted to MLX format from mpasila/Llama-3.1-Swallow-JP-EN-Translator-v1-8B. This model is specifically designed for Japanese-English translation tasks, leveraging the Llama 3.1 architecture. It offers a context length of 32768 tokens, making it suitable for handling longer translation inputs. Its primary strength lies in facilitating high-quality translation between Japanese and English.

Loading preview...

Overview

The moutons/Llama-3.1-Swallow-JP-EN-Translator-v1-8B-mlx-fp16 is an 8 billion parameter language model, specifically adapted for efficient deployment on Apple Silicon via the MLX framework. It is a converted version of the mpasila/Llama-3.1-Swallow-JP-EN-Translator-v1-8B model, utilizing mlx-lm version 0.31.2 for its conversion.

Key Capabilities

  • Japanese-English Translation: This model is fine-tuned and optimized for translating text between Japanese and English.
  • MLX Compatibility: Designed to run efficiently on Apple Silicon devices, leveraging the MLX framework for accelerated inference.
  • Llama 3.1 Architecture: Built upon the Llama 3.1 base, providing a robust foundation for language understanding and generation.
  • Extended Context Window: Supports a context length of 32768 tokens, allowing for the translation of longer passages and maintaining contextual coherence.

Good For

  • Developers working on applications requiring Japanese-English translation on MLX-compatible hardware.
  • Use cases where efficient, localized translation is critical.
  • Integrating translation capabilities into MLX-based projects and workflows.