ThomasComics/Quiller-AntiSlop-12B-v2

TEXT GENERATIONPricing:Input $0.87 / Cached $0.2 / Output $0.99Concurrent Unit Cost:1Model Size:12BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Dec 28, 2024Architecture:Transformer0.0K Featherless Exclusive Cold

ThomasComics/Quiller-AntiSlop-12B-v2 is a 12-billion parameter language model created by ThomasComics using the DARE TIES merge method. It combines elinas/Chronos-Gold-12B-1.0, TheDrummer/UnslopNemo-12B-v4.1, and DavidAU/MN-GRAND-Gutenberg-Lyra4-Lyra-12B-DARKNESS. This model is designed to leverage the strengths of its constituent models, offering a versatile foundation for various natural language processing tasks.

Loading preview...

Model Overview

ThomasComics/Quiller-AntiSlop-12B-v2 is a 12-billion parameter language model developed by ThomasComics. It was constructed using the DARE TIES merge method, a technique that combines multiple pre-trained models to create a new, potentially more robust model. The base model for this merge was elinas/Chronos-Gold-12B-1.0.

Merge Details

This model integrates components from two additional 12B parameter models:

The DARE TIES merge method, as described in the original paper and TIES-Merging paper, was applied with specific density and weight parameters for each contributing model. This approach aims to consolidate the diverse capabilities of the merged models into a single, cohesive unit.

Potential Use Cases

Given its merged nature, Quiller-AntiSlop-12B-v2 is suitable for a range of general-purpose NLP applications where a blend of different model strengths is beneficial. Its 12B parameter count and 32768 token context length provide a solid foundation for tasks requiring moderate complexity and context understanding.

Finetunes
2 models