Hermes-Instruct-7B-v0.2 Overview

Hermes-Instruct-7B-v0.2 is an instruction-tuned large language model developed by lodrick-the-lafted. It is built upon the Mistral-7B-Instruct-v0.2 base model and further fine-tuned using a selection from the teknium/openhermes dataset, specifically formatted in the Alpaca instruction style. This training approach aims to improve the model's ability to understand and follow diverse instructions.

Key Capabilities & Features

• Instruction Following: Enhanced through fine-tuning with Alpaca-formatted data, supporting both Mistral-Instruct and Alpaca prompt formats.
• Base Architecture: Leverages the efficient and capable Mistral-7B-Instruct-v0.2 model.
• Context Length: Supports a context window of 8192 tokens.
• Prompt Flexibility: Compatible with both <s>[INST] {sys_prompt} {instruction} [/INST] and Alpaca-style ### Instruction: {instruction} ### Response: prompting.

Performance Highlights

Evaluated on the Open LLM Leaderboard, Hermes-Instruct-7B-v0.2 achieved an average score of 63.82. Notable scores include:

• HellaSwag (10-Shot): 82.96
• Winogrande (5-shot): 76.87
• MMLU (5-Shot): 60.05
• AI2 Reasoning Challenge (25-Shot): 60.92

Recommended Use Cases

This model is suitable for a variety of instruction-based tasks, including but not limited to:

• General conversational agents
• Question answering
• Content generation based on specific prompts
• Educational tools requiring clear instruction adherence