yamatazen/Himeyuri-Magnum-12B-HereticMerge

TEXT GENERATIONConcurrent Unit Cost:1Model Size:12BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jul 12, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

Himeyuri-Magnum-12B-HereticMerge is a 12 billion parameter language model created by yamatazen, formed by merging yamatazen/Himeyuri-Magnum-12B and MuXodious/Mistral-Nemo-Instruct-2407-absolute-heresy using the SLERP method. This merged model aims to combine the strengths of its constituent models, offering enhanced performance for general language tasks. It is suitable for applications requiring a robust 12B parameter model derived from a strategic merge of existing instruction-tuned models.

Loading preview...

Model Overview

Himeyuri-Magnum-12B-HereticMerge is a 12 billion parameter language model developed by yamatazen. It is a product of a strategic merge operation, combining two distinct pre-trained language models to potentially leverage their individual strengths and improve overall performance.

Merge Details

This model was created using the SLERP (Spherical Linear Interpolation) merge method, a technique often employed to blend the weights of different models while preserving their learned representations. The merge process utilized mergekit, a tool designed for combining language models.

Constituent Models

The Himeyuri-Magnum-12B-HereticMerge is a blend of the following base models:

  • yamatazen/Himeyuri-Magnum-12B: One of the primary components contributing to the merged model's foundation.
  • MuXodious/Mistral-Nemo-Instruct-2407-absolute-heresy: The second key component, likely an instruction-tuned variant, contributing to the model's capabilities.

Configuration

The merge was performed with a t parameter of 0.5, indicating an equal weighting between the two merged models. The process used bfloat16 for both dtype and out_dtype, suggesting an optimization for memory and computational efficiency.

Potential Use Cases

Given its origin as a merge of instruction-tuned models, Himeyuri-Magnum-12B-HereticMerge is likely suitable for a variety of general-purpose language tasks, including:

  • Text generation
  • Instruction following
  • Question answering
  • Summarization