Riiid/sheep-duck-llama-2-13b
Riiid/sheep-duck-llama-2-13b is a 13 billion parameter language model with a 4096-token context length. This model is evaluated across various benchmarks, demonstrating capabilities in common sense reasoning, language understanding, and question answering. Its performance metrics indicate a general-purpose utility for a range of natural language processing tasks.
Loading preview...
Model Overview
Riiid/sheep-duck-llama-2-13b is a 13 billion parameter language model designed for general-purpose natural language understanding and generation tasks. It features a 4096-token context window, making it suitable for processing moderately long inputs.
Key Capabilities & Performance
Evaluated on the Open LLM Leaderboard, this model demonstrates a balanced performance across several key benchmarks:
- Common Sense Reasoning: Achieves 63.14% on ARC (25-shot) and 76.95% on Winogrande (5-shot).
- Language Understanding: Scores 84.52% on HellaSwag (10-shot) and 59.89% on MMLU (5-shot).
- Question Answering: Records 55.48% on TruthfulQA (0-shot).
- Mathematical Reasoning: Shows 9.17% on GSM8K (5-shot).
Use Cases
Given its broad performance across various benchmarks, Riiid/sheep-duck-llama-2-13b is suitable for applications requiring:
- General text generation and completion.
- Question answering systems.
- Reasoning tasks.
- Language understanding in diverse contexts.
Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.