Weyaxi/OpenHermes-2.5-Nebula-v2-7B

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Nov 12, 2023License:cc-by-nc-4.0Architecture:Transformer Open Weights Cold

OpenHermes-2.5-Nebula-v2-7B is a 7 billion parameter language model created by Weyaxi, resulting from a merge of teknium/OpenHermes-2.5-Mistral-7B and PulsarAI/Nebula-v2-7B-Lora. This model leverages the Mistral architecture, combining the strengths of its merged components to offer enhanced general-purpose language understanding and generation. It is designed for a broad range of applications requiring robust conversational AI and text completion capabilities.

Loading preview...

Overview

OpenHermes-2.5-Nebula-v2-7B is a 7 billion parameter language model developed by Weyaxi. It is a merged model, combining the foundational strengths of teknium/OpenHermes-2.5-Mistral-7B with PulsarAI/Nebula-v2-7B-Lora. This integration aims to leverage the distinct capabilities of both parent models to create a more versatile and performant language model.

Key Characteristics

  • Architecture: Based on the Mistral architecture, known for its efficiency and strong performance in its parameter class.
  • Merged Model: Benefits from the combined training and fine-tuning of two established models, potentially leading to improved generalization and specific task performance.

Performance

While specific benchmark scores are not detailed in the provided README, the model is listed on the Open LLM Leaderboard, indicating its participation in standardized evaluations. Users can refer to the leaderboard for up-to-date performance metrics across various tasks such as ARC, HellaSwag, MMLU, TruthfulQA, Winogrande, GSM8K, and DROP.

Potential Use Cases

  • General Text Generation: Suitable for a wide array of tasks including content creation, summarization, and creative writing.
  • Conversational AI: Can be applied in chatbots and virtual assistants due to its language understanding capabilities.
  • Research and Development: Provides a strong base for further fine-tuning on specific datasets or tasks.