djuna/MN-Chinofun-12B-4.1

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:12BQuant:FP8Ctx Length:32kArchitecture:Transformer0.0K Warm

djuna/MN-Chinofun-12B-4.1 is a 12 billion parameter language model created by djuna using the SCE merge method. It is based on ArliAI/Mistral-Nemo-12B-ArliAI-RPMax-v1.3 and integrates capabilities from five other pre-trained models. This model is designed to combine diverse strengths from its constituent models, offering a versatile foundation for various generative AI tasks.

Loading preview...

Model Overview

djuna/MN-Chinofun-12B-4.1 is a 12 billion parameter language model developed by djuna. This model was created using the SCE (Subspace Constrained Ensemble) merge method, a technique designed to combine the strengths of multiple pre-trained language models into a single, more capable model.

Merge Details

The base model for this merge is ArliAI/Mistral-Nemo-12B-ArliAI-RPMax-v1.3. The SCE method was applied to integrate the following five distinct models, aiming to leverage their individual characteristics:

  • Nitral-AI/Wayfarer_Eris_Noctis-12B
  • grimjim/magnum-twilight-12b
  • RozGrov/NemoDori-v0.2.2-12B-MN-ties
  • spow12/ChatWaifu_v1.4
  • GalrionSoftworks/Canidori-12B-v1

This approach suggests that MN-Chinofun-12B-4.1 is intended to inherit a broad range of capabilities from its diverse parent models, making it a potentially versatile option for general-purpose language generation and understanding tasks. The merge process utilized bfloat16 for its data type configuration.