giraffe176/Open_Hermes_Orca_Mistral-7B

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Feb 11, 2024License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

giraffe176/Open_Hermes_Orca_Mistral-7B is a 7 billion parameter language model created by giraffe176, merged using the task arithmetic method. It combines teknium/OpenHermes-2.5-Mistral-7B and Open-Orca/Mistral-7B-OpenOrca, leveraging their strengths. This model demonstrates strong performance across various reasoning and common sense benchmarks, making it suitable for general-purpose conversational AI and instruction-following tasks. It achieves an average score of 66.87 on the Open LLM Leaderboard.

Loading preview...

Model Overview

giraffe176/Open_Hermes_Orca_Mistral-7B is a 7 billion parameter language model developed by giraffe176. It was created using the task arithmetic merge method, combining two prominent Mistral-7B based models: teknium/OpenHermes-2.5-Mistral-7B and Open-Orca/Mistral-7B-OpenOrca. This merging approach aims to synthesize the capabilities of its constituent models.

Key Capabilities & Performance

This model demonstrates solid performance across a range of benchmarks, achieving an average score of 66.87 on the Open LLM Leaderboard. Specific benchmark results include:

  • AI2 Reasoning Challenge (25-Shot): 64.68
  • HellaSwag (10-Shot): 84.63
  • MMLU (5-Shot): 63.93
  • TruthfulQA (0-shot): 53.34
  • Winogrande (5-shot): 78.45
  • GSM8k (5-shot): 56.18

When to Use This Model

Given its strong performance in reasoning, common sense, and general knowledge tasks, this model is well-suited for:

  • General-purpose instruction following: Excelling in tasks where clear and accurate responses to prompts are required.
  • Conversational AI: Its merged nature from instruction-tuned models suggests good conversational abilities.
  • Reasoning tasks: Performance on ARC and MMLU indicates proficiency in logical deduction and understanding complex topics.