Lunzima/NQLSG-Qwen2.5-14B-MegaFusion-v8.8
TEXT GENERATIONConcurrency Cost:1Model Size:14.8BQuant:FP8Ctx Length:32kPublished:Mar 1, 2025Architecture:Transformer0.0K Cold

Lunzima/NQLSG-Qwen2.5-14B-MegaFusion-v8.8 is a 14.8 billion parameter language model created by Lunzima, merged from multiple Qwen2.5-14B variants using the SLERP method. This model combines characteristics from its constituent models, including NQLSG-Qwen2.5-14B-OriginalFusion, NQLSG-Qwen2.5-14B-MegaFusion-v8, and NQLSG-Qwen2.5-14B-MegaFusion-v5. It is designed to leverage the strengths of its merged components, offering a broad range of general-purpose language understanding and generation capabilities with a 32768 token context length.

Loading preview...