kekmodel/StopCarbon-10.7B-v3

TEXT GENERATIONConcurrency Cost:1Model Size:10.7BQuant:FP8Ctx Length:4kLicense:cc-by-nc-4.0Architecture:Transformer Open Weights Cold

kekmodel/StopCarbon-10.7B-v3 is an experimental 10.7 billion parameter language model created by kekmodel using mergekit, combining upstage/SOLAR-10.7B-Instruct-v1.0 and VAGOsolutions/SauerkrautLM-SOLAR-Instruct. This model leverages the TIES merging method to integrate the strengths of its base models, offering a unique instruction-following capability within a 4096-token context window. It is designed for general-purpose conversational AI and instruction-based tasks.

Loading preview...

kekmodel/StopCarbon-10.7B-v3 Overview

StopCarbon-10.7B-v3 is an experimental 10.7 billion parameter language model developed by kekmodel. It was created using the mergekit tool, specifically employing the ties merging method to combine two distinct base models:

  • upstage/SOLAR-10.7B-Instruct-v1.0: A robust instruction-tuned model.
  • VAGOsolutions/SauerkrautLM-SOLAR-Instruct: Another instruction-following variant based on the SOLAR architecture.

This merging approach aims to synthesize the capabilities and knowledge embedded within its constituent models, resulting in a unique instruction-tuned variant. The model operates with a context window of 4096 tokens.

Key Capabilities

  • Instruction Following: Designed to respond effectively to user instructions, leveraging the instruction-tuned nature of its base models.
  • General-Purpose Text Generation: Capable of generating coherent and contextually relevant text across a variety of prompts.
  • Merged Intelligence: Benefits from the combined strengths of SOLAR-10.7B-Instruct-v1.0 and SauerkrautLM-SOLAR-Instruct through the TIES merging technique.

Prompt Template

The model utilizes a specific prompt format for optimal interaction:

### User:
{user}

### Assistant:
{asistant}

Good For

  • Experimentation with merged models.
  • General instruction-based tasks and conversational AI applications.
  • Developers looking for a 10.7B parameter model with a unique lineage.

Finetunes
1 models