uukuguy/speechless-mistral-dolphin-orca-platypus-samantha-7b
The uukuguy/speechless-mistral-dolphin-orca-platypus-samantha-7b is a 7 billion parameter instruction-tuned language model, merged from ehartford/dolphin-2.1-mistral-7b, Open-Orca/Mistral-7B-OpenOrca, bhenrym14/mistral-7b-platypus-fp16, and ehartford/samantha-1.2-mistral-7b. Built on the Mistral-7B-v0.1 architecture, it features Grouped-Query Attention and Sliding-Window Attention, and demonstrates strong performance across various benchmarks, including an average score of 53.34 on the Open LLM Leaderboard. This model is designed for general-purpose conversational AI and instruction-following tasks, offering a robust base for diverse applications.
Loading preview...
Overview
uukuguy/speechless-mistral-dolphin-orca-platypus-samantha-7b is a 7 billion parameter instruction-tuned model, created by merging four distinct Mistral-7B based models: ehartford/dolphin-2.1-mistral-7b, Open-Orca/Mistral-7B-OpenOrca, bhenrym14/mistral-7b-platypus-fp16, and ehartford/samantha-1.2-mistral-7b. This unique combination aims to leverage the strengths of each component, building upon the Mistral-7B-v0.1 architecture which incorporates Grouped-Query Attention and Sliding-Window Attention for efficient processing.
Key Capabilities & Performance
- General Instruction Following: Designed for a broad range of conversational and instruction-based tasks.
- Benchmark Performance: Achieves an average score of 53.34 on the Open LLM Leaderboard, with notable scores including 64.33 on ARC, 84.4 on HellaSwag, and 63.72 on MMLU.
- Code Generation: Demonstrates a HumanEval score of 34.146, indicating its capability in Python code generation tasks.
Why Choose This Model?
This model is a strong candidate for developers seeking a versatile 7B parameter model that combines the fine-tuning benefits of multiple specialized Mistral-7B variants. Its balanced performance across various benchmarks makes it suitable for general-purpose applications where a blend of reasoning, common sense, and instruction-following is required.