ajtaltarabukin2022/merged_beat_champ_2model_ties

TEXT GENERATIONConcurrency Cost:2Model Size:32BQuant:FP8Ctx Length:32kPublished:Apr 18, 2026Architecture:Transformer Cold

The ajtaltarabukin2022/merged_beat_champ_2model_ties is a 32 billion parameter language model created by ajtaltarabukin2022 using the TIES merge method. This model combines pre-trained language models, specifically using dura-lori/affine-5DoKPQhZmKnFk4mNEmH4UorbqHDe3PFAPvEfJyDwNkimoAMe as a base and integrating RLStepone/Affine-h29-5Coip2NhkPhFCMLQ7LYs3zLVz9RSEZP7HJrakDeqM5RVdPs4. It is designed to leverage the strengths of its constituent models, offering a 32768 token context length for various language generation and understanding tasks.

Loading preview...

Model Overview

This model, merged_beat_champ_2model_ties, is a 32 billion parameter language model developed by ajtaltarabukin2022. It was constructed using the TIES (Trimming and Injecting Edges) merge method, a technique designed to combine the capabilities of multiple pre-trained language models efficiently. The model utilizes dura-lori/affine-5DoKPQhZmKnFk4mNEmH4UorbqHDe3PFAPvEfJyDwNkimoAMe as its foundational base.

Merge Details

The merging process involved combining dura-lori/affine-5DoKPQhZmKnFk4mNEmH4UorbqHDe3PFAPvEfJyDwNkimoAMe with RLStepone/Affine-h29-5Coip2NhkPhFCMLQ7LYs3zLVz9RSEZP7HJrakDeqM5RVdPs4. The configuration specified a bfloat16 data type and applied a weighted average to the model layers, with the base model contributing 60% and the additional model contributing 40% across layers 0 to 64. This method aims to create a more robust and capable model by integrating distinct features from its components.

Key Characteristics

  • Parameter Count: 32 billion parameters.
  • Context Length: Supports a substantial context window of 32768 tokens.
  • Merge Method: Employs the TIES method for combining pre-trained models.

Potential Use Cases

Given its large parameter count and extended context window, this model is suitable for:

  • Advanced text generation tasks requiring long-range coherence.
  • Complex language understanding and reasoning applications.
  • Scenarios benefiting from the combined knowledge and capabilities of its merged predecessors.