mavihsrr/GetCode-slerp

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Jan 15, 2024License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

GetCode-slerp is a 7 billion parameter language model created by mavihsrr, formed by merging CodeLlama-7b-Instruct-hf and codegen25-7b-multi using the slerp method. This model is specifically designed for code generation and instruction-following tasks, leveraging the strengths of its constituent models. It offers a 4096-token context length, making it suitable for various programming-related applications.

Loading preview...

Overview

GetCode-slerp is a 7 billion parameter model developed by mavihsrr, created through a merge of two prominent code-focused models: CodeLlama-7b-Instruct-hf and Salesforce/codegen25-7b-multi. This merge was performed using the slerp (spherical linear interpolation) method via LazyMergekit, aiming to combine and enhance their respective capabilities in code understanding and generation.

Key Capabilities

  • Code Generation: Inherits strong code generation abilities from both CodeLlama and codegen25.
  • Instruction Following: Designed to respond effectively to programming-related instructions.
  • Model Merging: Utilizes the slerp merge method, which can lead to a balanced integration of features from the base models.

Good For

  • Software Development: Assisting with writing and understanding code snippets.
  • Code Instruction Tasks: Generating code based on natural language prompts or completing code segments.
  • Experimentation: Users interested in merged models for specialized coding applications.