Casual-Autopsy/Maginum-Cydoms-24B
Casual-Autopsy/Maginum-Cydoms-24B is a 24 billion parameter language model with a 32768 token context length, created by Casual-Autopsy through a sophisticated merge of multiple pre-trained models. This model leverages TIES, DELLA, and SLERP merge methods to combine the strengths of various Mistral-based 24B models, including those from anthracite-core, TheDrummer, and zerofata. It is designed to offer enhanced performance by integrating diverse model capabilities, making it suitable for general-purpose text generation and understanding tasks.
Loading preview...
Maginum-Cydoms-24B: A Merged 24B Language Model
Maginum-Cydoms-24B is a 24 billion parameter language model developed by Casual-Autopsy, distinguished by its advanced merging methodology. This model integrates several pre-trained 24B models, primarily based on the Mistral architecture, to synthesize their collective strengths and improve overall performance.
Key Capabilities & Technical Details
- Advanced Merging Techniques: The model was constructed using a multi-stage merging process, incorporating sophisticated methods such as TIES, DELLA, and SLERP. This approach allows for a nuanced combination of different model characteristics.
- Diverse Model Integration: It merges contributions from various specialized 24B models, including
anthracite-core/Mistral-Small-3.2-24B-Instruct-2506-Text-Only,TheDrummer/Magidonia-24B-v4.3,TheDrummer/Precog-24B-v1,zerofata/MS3.2-PaintedFantasy-v3-24B,TheDrummer/Cydonia-24B-v4.3,ReadyArt/4.2.0-Broken-Tutu-24b, andzerofata/MS3.2-PaintedFantasy-v2-24B. - Context Length: The model supports a substantial context window of 32768 tokens, enabling it to process and generate longer, more coherent texts.
When to Use This Model
Maginum-Cydoms-24B is particularly well-suited for users seeking a robust, general-purpose language model that benefits from the combined intelligence of multiple specialized models. Its merged architecture aims to provide a balanced performance across a wide range of text-based tasks, making it a versatile choice for applications requiring strong language understanding and generation capabilities.