DreadPoor/Heart_Stolen-8B-Model_Stock
Heart_Stolen-8B-Model_Stock is an 8 billion parameter language model created by DreadPoor using the Model Stock merge method. It combines multiple Llama 3.1-based models and LoRAs, including Llama-3.1-Storm-8B, Llama-3.1-8B-Lexi-Uncensored-V2, NeverSleep/Lumimaid-v0.2-8B, DreadPoor/ScaduTorrent1.1-8b-model_stock, and mlabonne/Hermes-3-Llama-3.1-8B-lorablated, with a base of Sao10K/L3-8B-Stheno-v3.2 + ResplendentAI/Smarts_Llama3. This model is designed for general language tasks, leveraging the strengths of its diverse merged components to achieve a balanced performance across various benchmarks, with a context length of 8192 tokens.
Loading preview...
Overview
Heart_Stolen-8B-Model_Stock is an 8 billion parameter language model developed by DreadPoor. It was created using the Model Stock merge method, which combines several pre-trained language models and LoRAs to synthesize their capabilities. The merging process utilized mergekit and a base model composed of Sao10K/L3-8B-Stheno-v3.2 and ResplendentAI/Smarts_Llama3.
Key Capabilities
- Diverse Foundation: Built upon a foundation of multiple Llama 3.1-based models and specialized LoRAs, including akjindal53244/Llama-3.1-Storm-8B, Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2, NeverSleep/Lumimaid-v0.2-8B, DreadPoor/ScaduTorrent1.1-8b-model_stock, and mlabonne/Hermes-3-Llama-3.1-8B-lorablated.
- Balanced Performance: Achieves an average score of 28.98 on the Open LLM Leaderboard, with specific scores including 72.45 on IFEval (0-Shot) and 34.44 on BBH (3-Shot).
- General Purpose: Designed to handle a broad range of language understanding and generation tasks, benefiting from the varied expertise of its constituent models.
Good for
- General-purpose applications: Suitable for tasks requiring broad language capabilities due to its diverse merged components.
- Experimentation with merged models: Provides a practical example of the Model Stock merging technique for developers interested in model fusion.
- Applications requiring a balanced performance: Offers a solid baseline for various benchmarks, making it a versatile choice for different use cases.