Sorihon/Memorable-Dream-12B

TEXT GENERATIONConcurrency Cost:1Model Size:12BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Jun 23, 2026Architecture:Transformer0.0K Cold

Memorable-Dream-12B is a 12 billion parameter language model created by Sorihon through a Karcher Mean merge of seven distinct pre-trained models. This merge aims to combine the strengths of its constituent models, offering a versatile foundation for various natural language processing tasks. With its 12B parameter count, it provides a balance between performance and computational efficiency, suitable for applications requiring robust language understanding and generation.

Loading preview...

Memorable-Dream-12B: A Merged Language Model

Memorable-Dream-12B is a 12 billion parameter language model developed by Sorihon. It was constructed using the Karcher Mean merge method via mergekit, combining the capabilities of multiple pre-trained models.

Key Characteristics

  • Parameter Count: 12 billion parameters, offering a substantial capacity for complex language tasks.
  • Merge Method: Utilizes the Karcher Mean, a technique designed to effectively blend the weights of several models.
  • Constituent Models: The model is a composite of seven distinct base models, including:
    • DreadPoor/Irix-12B-Model_Stock
    • ReadyArt/Omega-Darker_The-Final-Directive-12B
    • Retreatcost/KansenSakura-Erosion-RP-12b
    • TheDrummer/Rocinante-X-12B-v1
    • Vortex5/Ethereal-Stardust-12B
    • Vortex5/Nether-Moon-12B
    • Vortex5/Stellar-Witch-12B

Potential Use Cases

Given its merged nature, Memorable-Dream-12B is designed to be a general-purpose language model. It can be adapted for various applications where a robust understanding and generation of text are required, potentially leveraging the diverse strengths inherited from its merged components. Developers seeking a versatile 12B model for experimentation or fine-tuning may find this model suitable.