Aryanne/MedWest-7B

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Apr 27, 2024License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

Aryanne/MedWest-7B is a 7 billion parameter language model created by Aryanne using the task_swapping merge method with mergekit. It is based on internistai/base-7b-v0.2 and incorporates senseable/WestLake-7B-v2. This model is primarily a research artifact for testing merge methods, offering a 4096 token context length.

Loading preview...

MedWest-7B Overview

MedWest-7B is a 7 billion parameter language model developed by Aryanne, created through the task_swapping merge method using mergekit. This model serves as a testbed for exploring advanced merging techniques, specifically utilizing internistai/base-7b-v0.2 as its foundational base and integrating senseable/WestLake-7B-v2.

Key Capabilities

  • Merge Method Exploration: Primarily designed for evaluating the task_swapping merge method.
  • Base Model Integration: Built upon internistai/base-7b-v0.2 for its core architecture.
  • Component Blending: Incorporates senseable/WestLake-7B-v2 to enhance its capabilities through merging.

Good for

  • Research and Development: Ideal for researchers and developers interested in experimenting with model merging techniques, particularly task_swapping.
  • Understanding Mergekit: Provides a practical example of a model created using the mergekit framework and its YAML configuration.
  • Comparative Analysis: Useful for comparing the performance and characteristics of models created via different merging strategies.