Edens-Gate/Control-8B-V1.1-abliteration

TEXT GENERATIONPricing:Input $0.2 / Cached $0.028 / Output $0.32Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Nov 7, 2024Architecture:Transformer0.0K Featherless Exclusive Cold

Edens-Gate/Control-8B-V1.1-abliteration is an 8 billion parameter language model created by Edens-Gate. It is a merge of Delta-Vector/Control-8B-V1.1 and grimjim/Llama-3-Instruct-abliteration-LoRA-8B, utilizing the task arithmetic method. This model is designed to combine the characteristics of its base models, offering a blended performance profile. With a context length of 32768 tokens, it is suitable for applications requiring moderate context understanding.

Loading preview...

Model Overview

Edens-Gate/Control-8B-V1.1-abliteration is an 8 billion parameter language model developed by Edens-Gate. This model was created using the mergekit tool, specifically employing the task arithmetic merge method. It combines the strengths of two distinct base models: Delta-Vector/Control-8B-V1.1 and grimjim/Llama-3-Instruct-abliteration-LoRA-8B.

Merge Details

The model's architecture is a direct result of merging its components with a weight of 1.0 for the combined base. This approach aims to integrate the learned representations from both source models into a single, cohesive model. The merge configuration specified bfloat16 as the data type and applied the task arithmetic method without normalization.

Potential Use Cases

Given its merged nature, this model is likely suitable for:

  • General text generation: Leveraging the capabilities inherited from its base models.
  • Exploratory AI development: For users interested in experimenting with merged model behaviors.
  • Applications requiring a blend of characteristics: Where the individual strengths of Control-8B-V1.1 and Llama-3-Instruct-abliteration-LoRA-8B are desired in a single package.

Finetunes
1 models