jlzhou/Qwen2.5-3B-Infinity-Instruct-0625
jlzhou/Qwen2.5-3B-Infinity-Instruct-0625 is a 3.1 billion parameter instruction-tuned causal language model developed by jlzhou. It is fine-tuned from Qwen/Qwen2.5-3B using a subset of the BAAI/Infinity-Instruct dataset. This model is designed for general instruction-following tasks, leveraging its base architecture and specific training data for conversational AI and response generation.
Loading preview...
Overview
jlzhou/Qwen2.5-3B-Infinity-Instruct-0625 is a 3.1 billion parameter instruction-tuned language model. It is built upon the Qwen/Qwen2.5-3B base model and has been fine-tuned using a specific subset (0625) of the BAAI/Infinity-Instruct dataset. This model is a result of distributed SFT (Supervised Fine-Tuning) techniques, as detailed in a related blog post.
Key Capabilities
- Instruction Following: Optimized for understanding and responding to user instructions.
- Conversational AI: Capable of generating coherent and contextually relevant responses in chat-like interactions.
- Base Model Enhancement: Leverages the capabilities of the Qwen2.5-3B architecture, known for its efficiency and performance in its size class.
Good For
- General Purpose Chatbots: Ideal for applications requiring a responsive and instruction-aware assistant.
- Prototyping and Development: Suitable for developers looking for a moderately sized, instruction-tuned model for various NLP tasks.
- Educational Tools: Can be used in scenarios where clear and concise explanations are needed based on prompts.
Performance Metrics
According to the Open LLM Leaderboard, the model achieves an average score of 16.61. Specific scores include 35.58 for IFEval (0-Shot), 26.91 for BBH (3-Shot), and 24.43 for MMLU-PRO (5-shot). These metrics provide an indication of its reasoning and general knowledge capabilities.