Nitral-AI/Wayfarer_Eris_Noctis-12B

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:12BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Jan 22, 2025Architecture:Transformer0.0K Warm

Nitral-AI/Wayfarer_Eris_Noctis-12B is a 12 billion parameter language model created by Nitral-AI, resulting from a slerp merge of Nitral-AI/Captain_Eris_Noctis-12B-alt-v0.420 and LatitudeGames/Wayfarer-12B. This model is designed for general text generation and comprehension tasks, leveraging its merged architecture to combine the strengths of its constituent models. With a 32768 token context length, it is suitable for applications requiring extensive contextual understanding and coherent long-form output.

Loading preview...

Model Overview

Nitral-AI/Wayfarer_Eris_Noctis-12B is a 12 billion parameter language model developed by Nitral-AI. It was created through a slerp merge of two distinct models: Nitral-AI/Captain_Eris_Noctis-12B-alt-v0.420 and LatitudeGames/Wayfarer-12B. This merging technique aims to combine the capabilities of both base models, offering a versatile solution for various language tasks.

Key Characteristics

  • Merged Architecture: Utilizes a slerp merge method, specifically blending layers from both source models with varying weights across self-attention and MLP components.
  • Parameter Count: A substantial 12 billion parameters, providing a balance between performance and computational requirements.
  • Context Length: Supports a context window of 32768 tokens, enabling the model to process and generate longer, more complex texts while maintaining coherence.
  • Prompt Format: Employs a ChatML-style prompt format, making it compatible with common instruction-following interfaces.

Good For

  • General Text Generation: Capable of producing diverse and coherent text outputs across a wide range of topics.
  • Context-Rich Applications: Its large context window makes it suitable for tasks requiring deep understanding of extensive input, such as summarization of long documents or complex conversational agents.
  • Exploration of Merged Models: Offers an opportunity to experiment with a model derived from a specific merging strategy, potentially yielding unique response characteristics.

Finetunes
1 models