Post
216
Meet LLaMAX2! Lightweight Pipeline - SFT on Qwen3-Instruct Models without Catastrophic Forgetting !!!
β¨Highlights:
πΉ SOTA Translation: State-of-the-art translation performance across both high- and low-resource trained languages.
πΉ Lightweight Pipeline: Engineered for efficiency, our pipeline uses minimal parallel data and applies layer-selective tuning to a powerful instruct model.
πΉ Strong Reasoning Capabilities: Exhibits reasoning abilities that are competitive with top-tier models like Qwen3-Instruct.
Welcome to use our models. More Details:
π Paper: LLaMAX2: Your Translation-Enhanced Model also Performs Well in Reasoning (2510.09189)
π Code: https://github.com/CONE-MT/LLaMAX2.0
π Model: LLaMAX/llamax20-68ad1c154fcf2623b75a068c
β¨Highlights:
πΉ SOTA Translation: State-of-the-art translation performance across both high- and low-resource trained languages.
πΉ Lightweight Pipeline: Engineered for efficiency, our pipeline uses minimal parallel data and applies layer-selective tuning to a powerful instruct model.
πΉ Strong Reasoning Capabilities: Exhibits reasoning abilities that are competitive with top-tier models like Qwen3-Instruct.
Welcome to use our models. More Details:
π Paper: LLaMAX2: Your Translation-Enhanced Model also Performs Well in Reasoning (2510.09189)
π Code: https://github.com/CONE-MT/LLaMAX2.0
π Model: LLaMAX/llamax20-68ad1c154fcf2623b75a068c