The ankitmaury7214/phi model is a 3.8 billion parameter, lightweight, instruction-tuned causal language model from the Phi-3 family, developed by Microsoft. It is optimized for strong reasoning, particularly in math and logic, and designed for memory/compute-constrained and latency-bound environments. This model supports a 4K token context length and excels in language understanding and code generation tasks among models under 13 billion parameters.
Loading preview...
Model Overview
This model, ankitmaury7214/phi, is a 3.8 billion parameter instruction-tuned variant of Microsoft's Phi-3-Mini-4K-Instruct. It is a lightweight, state-of-the-art open model trained on high-quality, reasoning-dense datasets, including synthetic data and filtered public web content. The model has undergone supervised fine-tuning and direct preference optimization to enhance instruction following and safety.
Key Capabilities
- Strong Reasoning: Excels in common sense, language understanding, math, code, and logical reasoning, performing comparably to larger models.
- Optimized Performance: Designed for environments with memory/compute constraints and latency-bound scenarios.
- Instruction Following: Improved instruction following and structured output capabilities, including explicit support for the
<|system|>tag in chat formats. - Code Generation: Demonstrates strong performance in code generation benchmarks like HumanEval and MBPP.
When to Use This Model
This model is ideal for commercial and research applications requiring a compact yet powerful language model. It is particularly well-suited for general-purpose AI systems where strong reasoning is critical, and computational resources are limited. While it shows lower performance in factual knowledge tasks like TriviaQA due to its size, this can be mitigated by augmenting it with external search capabilities.