DS-Archive/Chronos-Hermes-v2-13b-Limarp-Lora-Merged

TEXT GENERATIONConcurrency Cost:1Model Size:13BQuant:FP8Ctx Length:4kLicense:agpl-3.0Architecture:Transformer0.0K Open Weights Cold

DS-Archive/Chronos-Hermes-v2-13b-Limarp-Lora-Merged is a 13 billion parameter language model based on Llama 2 architecture, created by DS-Archive. This model is a merge of Chronos Hermes v2 13b and the LIMARP Lora adapter, specifically fine-tuned to enhance roleplay capabilities. It is optimized for generating diverse roleplay-oriented responses, differing from general-purpose LLMs by its specialized focus on conversational and character-driven interactions. The model has a context length of 4096 tokens and is intended for use in niche roleplaying applications.

Loading preview...

Model Overview

DS-Archive/Chronos-Hermes-v2-13b-Limarp-Lora-Merged is a 13 billion parameter model built upon the Llama 2 architecture. It is a merge of the Chronos Hermes v2 13b base model and the LIMARP Lora adapter, specifically updated in July 2023. The primary objective of this merge was to introduce a distinct roleplay "flavor" to the Chronos Hermes v2 model, making it suitable for character-driven interactions.

Key Capabilities

  • Enhanced Roleplay: Designed to provide varied and engaging responses for roleplaying scenarios.
  • Flexible Prompting: Supports two distinct instruction formats:
    • The Alpaca instruction format of the base model.
    • The LIMARP Lora instruction format, which includes <<SYSTEM>>, <<USER>>, and <<AIBOT>> tags for structured roleplay.

Limitations and Considerations

  • Bias: Exhibits biases similar to those found in niche online roleplaying communities, in addition to biases from its base model.
  • Factual Accuracy: Not intended for providing factual information or advice.
  • Training: As a merged model, its training details are derived from its constituent base model and Lora adapter.

When to Use This Model

  • Roleplaying Applications: Ideal for developers seeking a model specialized in generating creative and character-consistent responses for roleplay-focused use cases.
  • Conversational Agents: Suitable for building conversational agents where a specific persona or narrative style is desired.

Finetunes
1 models