Nitral-AI/Captain-Eris-Diogenes_Twilight-V0.420-12B is a 12 billion parameter merged language model, combining Nitral-AI/Captain-Eris_Twilight-V0.420-12B and Nitral-AI/Diogenes-12B-ChatMLified. This model leverages a slerp merge method with specific layer filtering to create a unique blend of capabilities. With a 32768 token context length, it is optimized for instruction-following and chat-based applications, offering enhanced conversational performance.
Loading preview...
Model Overview
Nitral-AI/Captain-Eris-Diogenes_Twilight-V0.420-12B is a 12 billion parameter language model created by Nitral-AI through a strategic merge of two distinct base models: Nitral-AI/Captain-Eris_Twilight-V0.420-12B and Nitral-AI/Diogenes-12B-ChatMLified.
Key Characteristics
- Merged Architecture: Utilizes a
slerp(spherical linear interpolation) merge method, applying specifictvalues to different layers (self_attn and mlp) to blend the strengths of its constituent models. - Parameter Count: Operates with 12 billion parameters, balancing performance with computational efficiency.
- Context Length: Features a substantial context window of 32768 tokens, enabling it to handle longer and more complex interactions.
- Instruction-Tuned: The inclusion of
Diogenes-12B-ChatMLifiedsuggests an emphasis on robust instruction following and chat-optimized responses.
Intended Use Cases
This model is particularly well-suited for applications requiring:
- Conversational AI: Its merged nature and ChatMLification indicate strong capabilities for engaging in dialogue and following conversational flows.
- Instruction Following: Designed to accurately interpret and execute user instructions, making it suitable for task-oriented applications.
- Extended Context Tasks: The large context window allows for processing and generating coherent text over longer inputs, beneficial for summarization, content generation, or complex query resolution.
Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.