arcee-ai/Saul-Instruct-Mistral-7B-Instruct-v0.2-Slerp

TEXT GENERATIONPricing:Input $0.4 / Cached $0.08 / Output $0.8Concurrent Unit Cost:1Model Size:7BQuant:FP8Context Size:4kTool Calling:SupportedPublished:Apr 16, 2024License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

arcee-ai/Saul-Instruct-Mistral-7B-Instruct-v0.2-Slerp is a 7 billion parameter instruction-tuned language model created by arcee-ai. It is a merged model, combining mistralai/Mistral-7B-Instruct-v0.2 and Equall/Saul-Instruct-v1 using the slerp merge method. This model leverages the strengths of its base components to offer enhanced instruction following capabilities within a 4096 token context window.

Loading preview...

Model Overview

arcee-ai/Saul-Instruct-Mistral-7B-Instruct-v0.2-Slerp is a 7 billion parameter instruction-tuned language model. It was created by arcee-ai through a merge operation using mergekit, combining two distinct models: mistralai/Mistral-7B-Instruct-v0.2 and Equall/Saul-Instruct-v1. This merging technique, specifically the slerp (spherical linear interpolation) method, aims to blend the characteristics and capabilities of the constituent models.

Key Characteristics

  • Merged Architecture: Built upon the Mistral-7B-Instruct-v0.2 base, enhanced by the Saul-Instruct-v1 model.
  • Slerp Merge Method: Utilizes spherical linear interpolation for combining model weights, with specific parameter adjustments for self-attention and MLP layers, as detailed in its configuration.
  • Instruction-Tuned: Designed to follow instructions effectively, inheriting this capability from its base models.

Good For

  • General Instruction Following: Suitable for a wide range of tasks requiring adherence to user prompts.
  • Experimentation with Merged Models: Provides a practical example of how different instruction-tuned models can be combined to potentially achieve synergistic performance.